luisa fernanda holguin colorado - qut › 98118 › 1 › luisa fernanda... · luisa fernanda...
TRANSCRIPT
IMPACT OF CONTACT LENS WEAR ON
CONJUNCTIVAL GOBLET CELLS
Luisa Fernanda Holguin Colorado
BSc (Optom)
Supervisors
Professor Nathan Efron
Dr Nicola Pritchard
Submitted in fulfilment of the requirements for the degree of
Doctor of Philosophy in Optometry
Institute of Health and Biomedical Innovation
Faculty of Health
School of Optometry and Vision Science
Queensland University of Technology
2016
Impact of Contact Lens Wear on Conjunctival Goblet Cells i
Keywords
Asymptomatic, conjunctiva, conjunctival impression cytology, contact lens, contact
lens wear, dry eye, goblet cells, goblet cell density, laser scanning confocal
microscopy, non-invasive, symptomatic
Impact of Contact Lens Wear on Conjunctival Goblet Cells ii
Abstract
Changes in conjunctival goblet cell density before and after contact lens wear and
associated with the development of dry eye are equivocal. In addition, contemporary
techniques such as in vivo laser scanning confocal microscopy to assess goblet cell
density, has not been compared with the traditional method of assessment ex vivo
conjunctival impression cytology. Through this investigation, valid and reliable
measurements of GCD were used in order to standardize and to inform the criteria of
the assessment techniques and also to advise future studies with regard to sampling
consistency and repeatability of measurements. The two methods used in this research
project to assess goblet cells were carefully described, and validation studies were
undertaken using immunohistochemical and immunocytochemical techniques.
Repeatability of measurements as well as the effect of test order were explored and an
image sample paradigm was also devised.
In vivo laser scanning confocal microscopy for goblet cell density assessment was
implemented as a less invasive and time-consuming technique and for the first time,
longitudinal assessment of goblet cell density using both in vivo and ex vivo
techniques is reported.
Contact lens wear induces a reduction of goblet cell density over 6 months, which is
exacerbated in those with dry eye symptoms. Either laser scanning confocal
microscopy or conjunctival impression cytology can be used to assess goblet cell
density in the conjunctiva. The link between the time course of change in conjunctival
goblet cell density and contact lens induced symptoms of dry eye is an important
contribution in the understanding process of contact lens discomfort, the major issue
related to contact lens wear discontinuation.
These findings are important for the development of methodologies for future
investigations of goblet cells assessment, specifically related to CL discomfort and
dry eye.
Impact of Contact Lens Wear on Conjunctival Goblet Cells iii
Table of Contents
Keywords .......................................................................................................................................... i
Abstract ............................................................................................................................................ ii
Table of Contents ............................................................................................................................ iii
List of Figures ................................................................................................................................. vi
List of Tables ................................................................................................................................ viii
List of Abbreviations ....................................................................................................................... ix
Statement of Original Authorship ...................................................................................................... x
Acknowledgements ......................................................................................................................... xi
INTRODUCTION .................................................................................................. 1 CHAPTER 1:
1.1 PREFACE ............................................................................................................................. 1
1.2 THESIS OUTLINE................................................................................................................ 3
1.3 SIGNIFICANCE OF THE STUDY ........................................................................................ 4
1.4 AIMS OF THE STUDY......................................................................................................... 5
1.5 RESEARCH QUESTIONS .................................................................................................... 5
1.6 INDIVIDUAL CONTRIBUTION TO THE RESEARCH TEAM ........................................... 6
LITERATURE REVIEW ...................................................................................... 7 CHAPTER 2:
2.1 THE OCULAR SURFACE .................................................................................................... 8 2.1.1 Tear Film .................................................................................................................... 8 2.1.2 The Cornea ................................................................................................................12 2.1.3 The Conjunctiva .........................................................................................................17
2.2 CONJUNCTIVAL GOBLET CELLS....................................................................................19 2.2.1 Development ..............................................................................................................20 2.2.2 Function.....................................................................................................................20 2.2.3 Proliferation ...............................................................................................................21 2.2.4 Secretion ....................................................................................................................21 2.2.5 Distribution ................................................................................................................22
2.3 GOBLET CELL ASSESSMENT TECHNIQUES..................................................................25 2.3.1 Biopsy .......................................................................................................................25 2.3.2 Conjunctival Impression Cytology ..............................................................................26 2.3.3 Laser Scanning Confocal Microscopy .........................................................................27 2.3.4 Correlation between CIC and LSCM ..........................................................................28
2.4 CONTACT LENSES ............................................................................................................28 2.4.1 History .......................................................................................................................28 2.4.2 Contact Lens-Induced Dry Eye Symptoms ..................................................................30
2.5 DRY EYE ............................................................................................................................31 2.5.1 Dry Eye Diagnosis, Criteria and Definition .................................................................31 2.5.2 Dry Eye Evaluation ....................................................................................................32
2.6 FACTORS AFFECTING GOBLET CELL DENSITY ..........................................................36 2.6.1 Factors influencing goblet cell differentiation .............................................................36 2.6.1 Ocular surface diseases affecting goblet cell density ...................................................36 2.6.2 The effect of contact lens wear on goblet cell densities ...............................................37
2.7 SUMMARY OF KNOWLEDGE GAPS AND OBJECTIVES OF RESEARCH PROGRAM .43
Impact of Contact Lens Wear on Conjunctival Goblet Cells iv
PRESUMED GOBLET CELLS ASSESSED BY LSCM CONFIRMED WITH CHAPTER 3:
IMMUNOHISTOCHEMISTRY IN A HUMAN PTERYGIUM BIOPSY ...................................47
3.1 PREFACE ............................................................................................................................47
3.2 PURPOSE ............................................................................................................................49
3.3 METHODS ..........................................................................................................................49
3.4 RESULTS ............................................................................................................................51
3.5 DISCUSSION ......................................................................................................................54
VALIDATION OF GIEMSA STAIN USING PAS FOR GOBLET CELL CHAPTER 4:
DENSITY ASSESSMENT .............................................................................................................57
4.1 PREFACE ............................................................................................................................57
4.2 PURPOSE ............................................................................................................................58
4.3 METHODS ..........................................................................................................................58
4.4 RESULTS ............................................................................................................................61
4.5 DISCUSSION ......................................................................................................................62
REPEATABILITY OF MEASURING GOBLET CELL DENSITY USING LSCMCHAPTER 5:
...............................................................................................................................65
5.1 PREFACE ............................................................................................................................65
5.2 PURPOSE ............................................................................................................................65
5.3 METHODS ..........................................................................................................................65
5.4 RESULTS ............................................................................................................................67
5.5 DISCUSSION ......................................................................................................................68
EFFECT OF TEST ORDER ON GOBLET CELLS ASSESSMENT ..................71 CHAPTER 6:
6.1 PREFACE ............................................................................................................................71
6.2 PURPOSE ............................................................................................................................72
6.3 METHODS ..........................................................................................................................72
6.4 RESULTS ............................................................................................................................74
6.5 DISCUSSION ......................................................................................................................75
IMAGE SAMPLE SIZE FOR GOBLET CELL DENSITY DETERMINATION77 CHAPTER 7:
7.1 PREFACE ............................................................................................................................77
7.2 PURPOSE ............................................................................................................................78
7.3 METHODS ..........................................................................................................................79 7.3.1 Conjunctival Laser Scanning Confocal Microscopy ....................................................79 7.3.2 Conjunctival Impression Cytology ..............................................................................84
7.4 DISCUSSION ......................................................................................................................89
GENERAL METHODOLOGY AND RESEARCH PLAN ..................................91 CHAPTER 8:
8.1 STUDY DESIGN .................................................................................................................91
8.2 RECRUITMENT ..................................................................................................................91
8.3 SAMPLE SIZE CALCULATION .........................................................................................91
8.4 STUDY POPULATION .......................................................................................................92 8.4.1 Inclusion Criteria .......................................................................................................95 8.4.2 Exclusion Criteria ......................................................................................................95
8.5 SCREENING AND BASELINE ...........................................................................................95
Impact of Contact Lens Wear on Conjunctival Goblet Cells v
8.6 DRY EYE AND OCULAR SURFACE ASSESSMENT ........................................................96 8.6.1 DEQ-5 Questionnaire .................................................................................................96 8.6.2 CLDEQ – 8 Questionnaire..........................................................................................97 8.6.3 Non-Invasive Break- Up Time Test ............................................................................97 8.6.4 Ocular Surface Staining ..............................................................................................98 8.6.5 Phenol Red Thread Test .............................................................................................98 8.6.6 Dry Eye and Ocular Surface Criteria for Enrolment ....................................................99
8.7 CONTACT LENS FITTING AND FOLLOW-UP VISITS ....................................................99
8.8 MASKING AND RANDOMIZATION ............................................................................... 100
8.9 RATIONALE ..................................................................................................................... 100 8.9.1 Rationale for 6-Month Study .................................................................................... 100 8.9.2 Rationale for Using Conventional Hydrogels ............................................................ 101
ASSOCIATION BETWEEN LSCM AND CIC FOR GOBLET CELL DENSITY CHAPTER 9:
ASSESSMENT ............................................................................................................................ 103
9.1 PREFACE .......................................................................................................................... 103
9.2 PURPOSE .......................................................................................................................... 104
9.3 METHODS ........................................................................................................................ 104
9.4 STATISTICAL ANALYSIS ............................................................................................... 105
9.5 RESULTS .......................................................................................................................... 105
9.6 DISCUSSION .................................................................................................................... 107
TIME COURSE OF CHANGES IN GCD IN SYMPTOMATIC AND CHAPTER 10:
ASYMPTOMATIC CONTACT LENS WEARERS ................................................................... 111
10.1 PREFACE .......................................................................................................................... 111
10.2 PURPOSE .......................................................................................................................... 111
10.3 METHODS ........................................................................................................................ 111
10.4 PARTICIPANTS ................................................................................................................ 111
10.5 STATISTICAL ANALYSIS ............................................................................................... 112
10.6 RESULTS .......................................................................................................................... 113
10.7 DISCUSSION .................................................................................................................... 119
SUMMARY, CONCLUSIONS AND RECOMMENDATIONS .................... 127 CHAPTER 11:
11.1 SUMMARY OF THE PROJECT OUTCOMES .................................................................. 127
11.2 CONTRIBUTION TO NEW KNOWLEDGE ...................................................................... 130
11.3 RECOMMENDATIONS FOR FUTURE RESEARCH........................................................ 130
Bibliography ................................................................................................................................. 133
Appendices.................................................................................................................................... 155 Appendix A: Ethics Clearance and Consent Form ................................................................ 155 Appendix B: Conference Presentations, Publications and Awards Arising from this
Thesis ...................................................................................................................... 163
Impact of Contact Lens Wear on Conjunctival Goblet Cells vi
List of Figures
Figure 2.1. Tear film structure. .......................................................................................................... 9
Figure 2.2. The corneal structure. Dua’s membrane is located between the stroma and Descemet
layers. .............................................................................................................................12
Figure 2.3. The conjunctiva structure. ..............................................................................................18
Figure 2.4. Goblet cell structure. ......................................................................................................20
Figure 2.5. Histogram shows the distribution of GCD values reported in the healthy bulbar
conjunctiva. The GCD values under the line represent 82% of 61 values reported. The
blue, red and green bars represent GCD of nasal and temporal, upper and lower
bulbar conjunctiva respectively. .......................................................................................24
Figure 3.1 The applanating surface of the TomCap in the horizontal position, the biopsy placed
on the centre of the TomCap. Excess liquid from the medium was not removed. ...............50
Figure 3.2. Immunolocalization of goblet cell markers in a pterygium biopsy (A) goblet cell
cytoplasm showed a band of CK-7 using FITC conjugated (green). (B) location of
mucin expression MUC5AC was labelled with TRITC (red). (C) 4, 6, diamidino-2-
phenylindole (DAPI) to identify nuclei (blue). (D) anti-mouse and anti-rabbit isotypes
control. Magnification 100X. ...........................................................................................52
Figure 3.3. Characterization of goblet cells from pterygium biopsy using laser scanning confocal
microscopy and immunohistochemistry. (A) in vivo LSCM image shows distinct
balloon-like cell appearance (yellow arrow) compared to the squamous non goblet
cells (dotted red arrow) (B) Immunofluorescence image of same tissue from image
(A) triple-labelled using FITC+TRITC+DAPI. The dotted arrow represents positive
stain for GC and solid arrow represents negative stain and presence of nucleus
assumed to be squamous non-goblet cells. (C) two times magnification from B
showing three distinct cell types with positive CK-7. Dashed arrow represents large
and oval-shaped GCs with positive MUC5AC expression (red). The dotted arrow
represents smaller and round-shaped GCs with less intensive red than the larger cells.
The solid line represents possibly immature GC lacking the balloon-like appearance and MUC5AC expression. ...............................................................................................53
Figure 4.1. Half of the CIC specimen stained using PAS (A) and the second half stained using
Giemsa (B). .....................................................................................................................60
Figure 4.2. Conjunctival impression cytology sample of nasal bulbar conjunctival stained with
PAS and counterstained with Gill’s haematoxylin stain. (magnification 200X). The
white arrow points to an epithelial cell and the black arrow points to a goblet cell. ............60
Figure 4.3. Bland-Altman plot of the relationship between differences in GCD obtained by
Giemsa and PAS staining procedures vs. GCD mean. The middle line represents the
mean difference between the two measurements (7 cells/mm²). The upper and lower
lines (dashed) represent the 95% LoA, +116 (upper bound) and -109 (lower bound)
including 0. There are 10 data points (1 per participant) that represent the difference
between Giemsa and PAS. Each data point represents the average value of GCD in 5 images at each testing time. One outlier is observed in the plot . .......................................62
Figure 5.1. In vivo confocal image of nasal bulbar conjunctival of goblet cells. The white arrow
points an epithelial cell and the blue arrow points a goblet cell. ........................................66
Figure 5.2. Bland-Altman plot of intra-observer test-retest of GCD. Relation between
differences in GCD vs. GCD mean. The middle line represents the mean difference
between the two measurements (14 cells/mm²). The upper and lower lines (dashed)
represent the 95% LoA, +191.87 (upper bound) and -162.41 (lower bound) including
0. There are 10 data points (1 per participant) that represent the difference between
test and retest. Each data point represents the average value of GCD in 3 images at
each testing time. .............................................................................................................68
Impact of Contact Lens Wear on Conjunctival Goblet Cells vii
Figure 6.1. Conjunctival impression cytology of nasal bulbar conjunctival stained with Giemsa
stain (magnification 200X). The white arrow points to an epithelial cell and the black
arrow points to a goblet cell. ............................................................................................74
Figure 7.1. Scatterplot of standard deviation vs. the number of images for 10 symptomatic CL
wearers. Image sampling analysis for LSCM. Each data point represents the standard
deviation of 3, 4, 5... and so on up to 30 images plotted against the number of images taken from 10 symptomatic contact lens wearers. A minimum of 11 random images
were necessary to determine the average of GCD. The average standard deviation
was approximately ± 40 cells/mm². ..................................................................................84
Figure 7.2. Photograph of the eye under examination using slit lamp and fluorescein conducted
under cobalt blue illumination with a yellow filter. Location of the impression and
evaluation of the integrity of the exposed bulbar conjunctiva is observed. .........................85
Figure 7.3. Staining procedure using Giemsa stain. (A) More than 60% of the filter is covered in
cell material. (B) Staining procedure in a well plate sample holder. (C) The filter of
the Millicell cell culture insert detached from the plastic ring and ready for imaging
onto a slide and covered with a coverslip. ........................................................................86
Figure 7.4. Scatterplot of standard deviation vs. the number of images for 10 symptomatic CL
wearers. Image sampling analysis for CIC. Each data point represents the standard deviation of 3, 4, 5... and so on up to 10 images plotted the against number of images
taken from 10 symptomatic contact lens wearers. A minimum of 5 images was
necessary to determine the average of GCD. The average of the standard deviation
was approximately ± 152 cells/mm² .................................................................................88
Figure 8.1. The bar chart represents the country of origin in percentage of the study population
that was examined at baseline (N = 83). ...........................................................................93
Figure 8.2. A schematic representation of the study population enrolled, excluded, the number
of examinations at baseline and those assigned to DE groups after the 1-week visit. ..........94
Figure 8.3. Efron grading scale system for ocular surface assessment. ..............................................98
Figure 9.1. Bland-Altman plot of the differences in GCD between the two techniques versus the
average. The middle, heavier line represents the linear regression and the lighter upper and lower lines represent the 95% LoA. Each of 90 data points represents the
average value of 5 and 11 images with CIC and LSCM, respectively. LSCM and CIC
agree at any time between 48.52 to 213.31 (upper bound) and -0.43 to -0.08 (lower
bound). .......................................................................................................................... 106
Figure 10.1. Distribution and number of participants examined at four visits. .................................. 113
Figure 10.2. Line graph of the longitudinal course of goblet cell density over a 6-month period
in CL wearers and controls assessed with (A) laser scanning confocal microscopy and
(B) conjunctival impression cytology. On each graph, the green line represents the
asymptomatic CL-induced DE group, the red line the symptomatic CL-induced DE
group and the blue line the control participants. Error bars indicate mean ± SD. .............. 116
Figure 10.3. Line graph of the longitudinal course of goblet cell density over a 6-month period,
excluding the 1-week visit, in CL wearers and controls assessed with (A) laser scanning confocal microscopy and (B) conjunctival impression cytology. On each
graph the green line represents the asymptomatic CL-induced DE group, the red line
symptomatic CL-induced DE group, and the blue line the control participants. Error
bars indicate mean ± SD. ............................................................................................... 117
Impact of Contact Lens Wear on Conjunctival Goblet Cells viii
List of Tables
Table 1.1. Responsibilities of examiners............................................................................................ 6
Table 2.1. Goblet cell density assessed by conjunctival impression cytology in contact lens
wear; ordered by direction of change................................................................................42
Table 4.1. Basic descriptive statistics for GCD collected by CIC technique using Giemsa and PAS staining procedures. Values are presented in mean ± SD or count for categorical
variables. .........................................................................................................................61
Table 5.1. Demographic characteristics of participants of the intra-observer repeatability
analysis of GCD measures using LSCM. Values are presented in mean ± SD or count
for categorical variables. ..................................................................................................67
Table 5.2. Mean difference, ICC and LoA of the intra-observer repeatability analysis of GCD
measures using LSCM. ....................................................................................................67
Table 6.1. Descriptive statistics for GCD assessed by CIC with and without prior LSCM and
paired sample t-test between the two main outcome variables (OD = LSCM prior
GCD (CIC)) and (OS = GCD (CIC)) Values are presented as mean ± SD. ........................74
Table 7.1. Previous observations of goblet cells under LSCM. ..........................................................80
Table 8.1. DE and ocular surface assessment cut-point indicative of CL-induced DE symptoms........99
Table 9.1. Ocular surface integrity and dry eye assessment of participants ...................................... 106
Table 10.1. Demographic and clinical characteristics of the participants at the baseline and after
6 months of contact lens wear. ....................................................................................... 114
Table 10.2. The effect of time (visit), group and test on the dependent variable GCD assessed
with LSCM using the type III of fixed effects from LMM analysis. ................................ 118
Table 10.3. Maximum likelihood of the fixed effect parameters for LMM with GCD assessed
with LSCM as the continuous response variable. ............................................................ 119
Impact of Contact Lens Wear on Conjunctival Goblet Cells ix
List of Abbreviations
mRNA
CIC
CK-7
CLDEQ-8
CL(s)
CLW
DAPI
DE
DEQ-5
DEWS
DMEM
DNA
EGF
FC
FITC
GC(s)
GCD
HI-FCS
HLA-DR
HRT III
ICC
IgG
IHBI
IHC
LMM
LoA
LSCM
MUC(s)
NIBUT
OD
OS
OSS
PAS
PBS
PCR
PMMA
PRT
QUT
RGP
SD
SEM
TBUT
TFOS
TRITC
messenger ribonucleic acid
Conjunctival impression cytology
Cytokeratin 7
Contact lens dry eye questionnaire-8
Contact lens(es)
Contact lens wear
Diamidino-2-phenylindole dihydrochloride
Dry eye
Dry eye questionnaire-5
Dry eye workshop
Dulbecco's modified eagle's medium
Deoxyribonucleic acid
Epidermal growth factor
Flow cytometry
Fluorescein isothiocyanate
Goblet cell(s)
Goblet cell density
Heat-inactivated foetal calf serum
Human leukocyte antigen d-related
Heidelberg Retinal Tomograph III
Interclass correlation coefficient
Immunoglobulin G
Institute of Health and Biomedical Innovation
Immunohistochemistry
Linear mixed model
Limits of agreement
Laser scanning confocal microscope/microscopy
Mucin(s)
Non-invasive tear break-up time
Oculus dexter (right eye)
Oculus sinister (left eye)
Ocular surface staining
Periodic acid-Schiff
Phosphate-buffered saline
Polymerase chain reaction
Polymethylmethacrylate
Phenol red thread
Queensland University of Technology
Rigid gas permeable
Standard deviation
Standard error of the mean
Tear break-up time
Tear film and ocular surface society
Tetramethylrhodamine
Impact of Contact Lens Wear on Conjunctival Goblet Cells x
Statement of Original Authorship
The work contained in this thesis has not been previously submitted to meet
requirements for an award at this or any other higher education institution. To the best
of my knowledge and belief, the thesis contains no material previously published or
written by another person except where due reference is made.
Signature:
Date: 8th
August 2016
_________________________
QUT Verified Signature
Impact of Contact Lens Wear on Conjunctival Goblet Cells xi
Acknowledgements
This project was funded in part by a grant from the Cornea and Contact Lens Society
of Australia and undertaken with an Australian Postgraduate Award. In kind support
Coopervision Australia donated the contact lenses used in this study.
I acknowledge my supervisors Professor Nathan Efron and Dr Nicola Pritchard for
their guidance and support during my PhD candidature. I must thank Miss Emily
Bryan and Miss Maria Lourdes Muerza-Cascante for their laboratory advice and Dr
Cirous Dehghani for the statistical support.
Deep thanks go to the people who encourage me emotionally throughout this journey
and to all the participants who have volunteered for this study.
Luisa H. Colorado
Introduction 1
Introduction Chapter 1:
1.1 PREFACE
Contact lenses are medical devices mainly used for visual improvement. Cosmetic,
therapeutic and practicality are factors that can motivate people to wear contact lenses
instead of spectacles. Data from 2004 showed that approximately 125 million people
(2%) of the worldwide population wear contact lenses (Barr, 2005). The 2014 Annual
Report on Contact Lenses showed an increase of approximately 1.2 million new
contact lens wearers in the United States alone from 2004 to 2014 (Nichols, 2015).
However, the rate of contact lens discontinuation has remained constant over the past
few years, with the cause of discontinuation being ocular dryness among others
(Pritchard et al., 1999).
A good quality tear film is important to the health of the eye because it protects the
ocular surface against chemical, physical, and microbiological injuries. The tear film
has manly three components: lipids, water, and mucins. The mucins are important
because they provide lubrication, increases water retention, and create a barrier to
infectious agents. Approximately 90% of these mucins are produced by conjunctival
goblet cells.
The presence of mucin during contact lens wear promotes surface wettability and
tolerability of the contact lens on the eye. However, a lack of mucin discharge onto
the ocular surface can cause the patient to experience symptoms such as itching,
burning, dryness and reduction of visual acuity (Hori et al., 2006).
The symptom of dryness during contact lens wear is not completely understood but,
nonetheless, is a very common clinical issue in contact lens wearers. Dryness may
cause contact lens wearers to reduce contact lens wear, or cease lens wear altogether.
It has been shown that approximately 50% of soft contact lens wearers develop dry
eye symptoms (Pritchard et al., 1999).(Pritchard et al., 1999)
The results of previous investigations into the impact of contact lens wear on
conjunctival goblet cell density are equivocal, perhaps as a result of methodological
Introduction 2
differences in using conjunctival impression cytology to determine goblet cell density.
Conjunctival impression cytology has being used in association with flow cytometry
and polymerase chain reaction. These supplementary procedures enable the analysis
of mucin-type material and allow researchers to quantify the amount of mucin
produced by goblet cells. Polymerase chain reaction is a modern technology capable
to identify, cloning, and characterising the mucin gene (MUCs) and facilitates goblet
cell quantification based on their specific mucin type MUC5AC.
Due to the invasive nature of the technique, time-intensive and laboratory-based work
required for conjunctival impression cytology, in vivo laser scanning confocal
microscopy may represent an alternative and fundamentally advantageous approach to
the morphological assessment for investigation of the ocular surface. Therefore, it is
important to establish whether there is an association between the non-invasive in
vivo and moderately invasive ex vivo techniques for assessment of goblet cells,
especially in contact lens wear. Hence, part of this study was design to determine
whether the in vivo method of measuring goblet cell density could be used in place of
conjunctival impression cytology.
The time-course of changes to goblet cell density as a result of contact lens wear is
still unclear, and there is, to my knowledge, no evidence of longitudinal studies of
goblet cell density assessment using in vivo laser scanning confocal microscopy in
contact lens wear. The work conducted to date, however, suggests contact lens wear
does alter normal goblet cell density in some way. It is therefore important to
understand whether contact lens wear causes a reduction in goblet cell density in non-
contact lens wearers who are introduced to contact lens wear and to determine the link
between goblet cell density and dry eye symptoms related to contact lens wear.
This work could help researchers and practitioners to understand the importance of
these cells and the role they play in the comfort of contact lens wear. Longitudinal
data relating goblet cell density and contact lens-induced dry eye symptoms are
lacking, along with goblet cell density assessed over time in contact lens wear using
in vivo laser scanning confocal microscopy.
Further, the findings of this study would help develop a standardized methodology for
understanding the variation in goblet cell density.
Introduction 3
1.2 THESIS OUTLINE
This thesis is presented for the PhD by monograph and contains studies that attempted
to resolve discrepancies in the literature concerning the time course of changes of
GCD in CLW using validated methodological procedures. The overall outline of this
thesis contains 11 chapters, including the introduction chapter.
In Chapter 2 the literature review was orientated to the previous studies looking at the
importance of GCD in the human eye and the effect of CLW. In this Chapter the
techniques of assessments and the DE symptoms in CL wearers are described.
In Chapter 3 the presumed GCs assessed by LSCM were confirmed with
immunohistochemistry (IHC) in a human pterygium biopsy. This Chapter reports the
characterization of GCs morphologically identified with LSCM by IHC using
antibodies for mucins such as MUC5AC.
In Chapter 4 the validation of Giemsa stain procedure using periodic acid-Schiff
(PAS) reagent for mucus cell detection was determined. In this Chapter, the validation
of Giemsa stain is demonstrated by comparing samples from same participants using
PAS. The importance of this study was to demonstrate an association between cell
counts from CIC specimens using Giemsa and PAS staining techniques in order to
determine if Giemsa stain is an accurate and valid technique for the assessment of
GCD.
In Chapter 5 the repeatability of measuring GCs using LSCM was demonstrated. This
study explores the intra-observer repeatability in assessing GCs using LSCM at two
different times in the same participants. Satisfactory repeatability is important in
measurement analysis because it detects critical differences in monitored values. The
Bland-Altman repeatability was plotted and inter-class correlation coefficients
determined for one observer measuring GCD on two separate occasions.
In Chapter 6 the effect of test order on GCs assessment was reported. In this study, the
hypothesis relates the degree of differences between two main outcome variables
which are LSCM prior CIC. The correlation and the similarities between these two
measurements are established in this Chapter.
Introduction 4
In Chapter 7 the image sample size for GCD using LSCM and CIC was determined.
This analysis explores an optimal image sampling for estimation of GCD using
LSCM and CIC. The objective of this experiment was to determine the number of
images required to provide an accurate estimate of GCD for each technique of
assessment.
In Chapter 8 the general methodology and research plan are explained. This Chapter
describes the study design, participant cohort, and the assessment techniques
The following Chapters describe the main outcomes of the investigation.
In Chapter 9 the association between LSCM and CIC for the assessment of
conjunctival GCs is stablished. The strength of relationship between LSCM and CIC
techniques for GCs assessment is described in this Chapter.
In Chapter 10 the time course of changes in GCD in symptomatic and asymptomatic
CL wearers is demonstrated. This Chapter documents the longitudinal assessment of
GCD over a 6-month period and also highlights the interaction of LSCM and CIC
over time.
Concluding remarks are presented in Chapter 11 and include study logistics such as
assessment of the GCs using LSCM and CIC carried out by two different observers.
The LSCM images from which GC counts were determined was captured by observer
one and the assessment of GCD was calculated by observer two.
1.3 SIGNIFICANCE OF THE STUDY
Real-time observations of cell layers from the ocular surface using LSCM have been
introduced as a useful measurement for assessing GCD. Previous cross-sectional
studies have suggested that LSCM is a non-invasive tool to monitor changes of the
ocular surface that uses a through-focusing technique allowing the visualisation of the
entire thickness of the conjunctive. Based on the recent report from the Tear Film and
Ocular Surface Society (TFOS), there is a lack of evidence to support the association
of GCs with CL discomfort. Hence, by determining longitudinal changes of GCD
related to CL-induced DE, GCs could be adopted as a marker for future studies
related to CL discomfort. Additionally, this investigation provides, for the first time,
Introduction 5
evidence related to the capability of LSCM as a repeatable technique and also
establishes its reliability for GC assessment. It was demonstrated in this work, by
employing immunohistochemical staining techniques on a biosample of human
conjunctival tissue, that the cells presumed to be goblet cells observed using LSCM
are likely to be goblet cells.
1.4 AIMS OF THE STUDY
The aim of this research was to use both LSCM and CIC to assess GCD over time and
defined the influence of CL-induced DE on GCD. Participants were recruited into
conventional hydrogel daily disposable CLs and monitored over a 6-month period. To
determine the impact of CL-induced DE, the CL group was subdivided into those with
symptomatic DE induced by CLW and those without DE symptoms; a concurrent
non-lens-wearing control group was assessed.
The overall aims of this study were:
• To determine the time course of changes in GCD in symptomatic and
asymptomatic CL wearers in those with and without CL-induced DE using
LSCM and CIC.
• To determine the association between in vivo LSCM and ex vivo CIC as an
assessment technique for conjunctival GCD.
1.5 RESEARCH QUESTIONS
From the primary aims of this study, the following research questions and hypotheses
were posed:
1. Is GCD different in individuals who develop symptoms of DE from CLW
compared to those who do not develop symptoms?
Hypothesis: There is a reduction of GCD in symptomatic participants for
DE induced by CLW compared to asymptomatic CLW and a control group.
2. Are the in vivo and ex vivo techniques of measuring GCD in CLW
equivalent?
Introduction 6
Hypothesis: There is a strong association between techniques over time, thus
the less invasive measure can be used in future research.
1.6 INDIVIDUAL CONTRIBUTION TO THE RESEARCH TEAM
The research was conducted alongside a study undertaken by another PhD student,
Yahya Alzahrani, at the Institute of Health and Biomedical Innovation (IHBI),
Queensland University of Technology (QUT). These two studies utilised the same
participants and applied a similar methodology. Individual research programs and
questions were developed which address different tissue changes associated with
CLW which required diverse independent assessment techniques. The duties of the 6-
month clinical investigation were shared between both candidates to assist with
masking and labour distribution as shown in Table 1.1, and analysis of variables was
conducted by each candidate independently.
Table 1.1. Responsibilities of examiners.
Luisa Holguin Yahya Alzahrani
Preliminary screening (history, lensometry, visual acuity, subjective refraction, slit lamp exam)
DE and ocular surface assessments CIC
CL slit-lamp examination LSCM CL fitting CL follow-up
Literature Review 7
Literature Review Chapter 2:
People may choose to wear CLs in order to improve their quality of life, to facilitate
athletic participation, or to improve their peripheral vision. CLs can also be used for
therapeutic purposes such as drug delivery and as bandages for superficial eye injuries
or post-surgically. Specially designed lenses can change the shape of the eye to
correct short-sightedness as well.
A good-quality tear film is important to the health of the eye because it protects the
ocular surface against chemical, physical, and microbiological insult. The tear film is
composed of three layers: mucin, aqueous, and lipid. The mucin layer is important
because it provides lubrication, increases water retention, and creates a barrier to
infectious agents. The mucin layer is produced primarily by conjunctival GCs.
The presence of mucin during CLW promotes surface wettability and tolerance of the
CL by the eye. When there is a reduced level of mucin on the ocular surface, the
patient may experience symptoms such as itching, burning, dryness and reduction of
visual acuity.
The results of previous investigations into the impact of CLW on conjunctival GCD
are equivocal, perhaps as a result of the methodological difficulties of CIC to
determine GCD. There are two approaches to GC assessment: the established method
of CIC and a newer, non-validated method, in vivo LSCM.
In order to address the relationship between CLW and GCs, it is important to
understand CL interaction with the superficial ocular structures, especially with the
conjunctiva and tear film, the techniques of GC assessment and the DE symptoms
induced by CLW and reductions of GCD. These CL interactions and techniques of
GC assessment are discussed in the following sections.
This Chapter begins with a review of the literature regarding the anatomical
components of the ocular surface that interact with the CL when it is placed into the
eye such as the tear film, the cornea and the conjunctiva including GCs.
Literature Review 8
2.1 THE OCULAR SURFACE
The ocular surface is formed by the continuity of the corneal and the conjunctival
epithelium joined by the limbus and its mucosal adnexa to the accessory lacrimal
glands, excretory ducts and the lacrimal drainage system including the lacrimal
canaliculi (Knop et al., 2010). Furthermore, these epithelia have the same
embryological derivation (Ridley, 1963). This broader concept, with some additional
features, has been called the ocular surface system and broadly comprises the cornea,
the conjunctiva, and the lacrimal and accessory meibomian glands (Stern et al., 1998).
The ocular surface system main function is to maintain the optical properties of the
cornea and to withstand physical and chemical trauma as well as to protect the eye
against internal and external changes that may affect the physiology of the ocular
surface (Rolando and Zierhut, 2001). CLs in the eye can potentially induce changes in
the dynamic of the ocular surface (Pisella et al., 2001).
2.1.1 Tear Film
The tear film is the fluid that covers the ocular surface. This fluid is formed by the
tear menisci and the precorneal tear film; the latter has a higher interaction with CLW.
The precorneal tear film is approximately 3 µm thick when the eye is open, its volume
is about 6.2 µl (Nichols et al., 2002) and its maximum thickness is estimated to be
about 3 to 4 μm. The tear film is thicker immediately after eyelids open and then starts
to decrease immediately before blinking (King-Smith et al., 2004).
Literature Review 9
Figure 2.1. Tear film structure.
Structural layers
The tear film is not a homogeneous liquid phase but rather structured in distinct layers
that are roughly parallel to the corneal and conjunctival surface. The tear film is a
dynamic changing structure with a complex combination of components that originate
in different parts of the eyelids and the eyeball (Doane and Gleason, 1994).
The first researcher to suggest that three layers formed the tear film structure was
Wolff (1946). This concept involved an anterior lipid layer, an intermediate aqueous
and a posterior mucus layer. Tiffany and Bron (1978) suggested that the tear film
structure was formed by five layers; an anterior oily layer, a polar lipid layer, a
mucosal layer, an aqueous layer and the glycocalyx layer of the corneal surface which
is a mucopolysaccharide component of low molecular weight capable of producing
hydrophilicity in the corneal epithelium. In 1984, a new theory emerged that only two
layers formed the tear film; one superficial lipid layer and one glycoproteic aqueous
layer (Holly, 1984). Observations by Dilly (1985) showed that there is a high number
of elongate vesicles at the anterior pole of the corneal and conjunctival epithelium.
These intracellular vesicles float in tear film and release mucins onto the ocular
surface. Some other authors have speculated that there is an independent layer in the
Literature Review 10
tear film called the mucin-aqueous complex which is a phase between the hydrophilic
substrate of the mucin layer and the hydrophobic surface of the outer layer of the tear
film (Chen et al., 1997; Price-Schiavi et al., 1998).
The latest evidence on tear film structure model will be used to define tear film
structure in this thesis.
The lipid layer
The lipid layer is a thin anterior phase of the tear film located between the air and the
mucin-aqueous complex. It is approximately 0.1 µm thick and is produced by the
sebaceous secretion of the meibomian glands (Linton et al., 1961). In fact, polar lipids
from acinar cell secretion are also present in the lipid layer (Ehlers, 1965). However,
in CLW thinner measurements of this layer have been reported (Guillon et al., 1997).
The lipid layer is composed of two phases; one is a thin internal hydrophilic layer
with positive and negative charges with polar lipids such as phospholipids and
ceramides and a second thick external hydrophobic phase which is in contact with the
air and has anti-evaporative properties that contain waxes, cholesterol esters,
triglycerides and fatty acids (Bron et al., 2004).
Evaluation of the lipid layer of the tear film margin is called meibometry. This
technique was developed to measure the levels of lipid in the lower lid margin. It
involves blotting of the lower central lid margin onto a plastic tape and reading of the
amount taken up by optical densiometry. This provides an indirect measure of the
meibomian lipid levels, which are in the range of 300 µm in healthy adults (Chew et
al., 1993). Another technique to evaluate the thickness of the lipid layer is called
interferometry. The principle of this measurement is to assess the tear film by using
specular reflection of white light using interferometry colour units, which indicate the
index of the thickness based on the estimation of the mean interference of the colours
(Guillon, 2002).
The mucin-aqueous complex
The mucin-aqueous complex of the tear film is an abundant hydrated glycoproteins
and an isotonic aqueous solution composed of antibacterial properties such as
Literature Review 11
lysozyme and lactoferrin. It also carries albumin, lipocalin, fibroblast growth factor,
nerve growth factor, albumins A, G, M, E among others, glucose, glycogen and
organic salts that provide nutrients to the cornea (Van Haeringen, 1981). This phase is
composed mainly of the secretion of the main lacrimal and accessory glands, i.e.
Krause and Wolfring glands. The lacrimal gland also secretes a variety of cytokines
(small signalling proteins) including interleukin-1, tumour necrosis factor, epidermal
growth factor (EGF) and transformed growth factor beta-1. It has been suggested that
these cytokines are important to maintain homoeostasis and ensure the health of the
ocular surface (Pflugfelder, 1998).
The mucous phase is a thin glycoprotein layer, highly hydrated, which covers the
corneal and conjunctival epithelium over the glycocalyx.
The term mucin refers specifically to molecules of protein that contain
glycoconjugates; in contrast, mucus is a compound of mucins, inorganic salts,
desquamated cells and leukocytes and mucous is its adjectival form (Dorland, 2012).
Previously it was thought that the glycoproteins produced by the epithelial cells from
the cornea and the conjunctiva were acidic and belonged to the mucous layer and
those produced by the lacrimal gland were neutral and sparse in the aqueous layer
(Report of the International Dry Eye WorkShop (DEWS), 2007). Recent genetic
studies have identified 17 types of mucin in the human epithelium, and they have
been classified according to their function and origin; gel-forming/or secretory and
membrane-associated. In the ocular surface mucins expressed by the GCs and the
lacrimal gland (MUC5AC and MUC7, respectively) are known to be gel-forming/or
secretory mucins. Additionally, mucins MUC 1, 4 and 16 are the membrane-
associated type, and they are expressed by the corneal and conjunctival epithelium
(Gipson, 2004).
The transition of ocular surface and the tear film is composed of hydrated
glycoproteins that are secreted by the GCs and the membrane-associated mucins
MUC1 and MUC4. The union of these mucins creates a hydrophilic substrate that
leads to the transition of the hydrophobic mucins to the hydrophilic surface of the
aqueous solution. The mucin released by GCs (MUC5AC) is the main component of
the ocular mucin gel which keeps the tear film steady during blinking (Holly and
Literature Review 12
Lemp, 1971). In rigid CL wearers, a decrease in mucus production (Pisella et al.,
2001) have been demonstrated.
The resistance of the tear film to gravity is simply due to surface forces such as
viscosity and gels of the tear film (Holly, 2005). However; there is no evidence under
physiological conditions of such assumptions.
2.1.2 The Cornea
Structural layers
The cornea is a highly differentiated tissue that allows refraction and transmission of
light to the retina. The corneal shape is that of a concave-convex lens with a front
surface in contact with the pre-corneal tear film and a back surface coated with
aqueous fluid. These fluids are responsible for maintenance of the corneal
transparency and shape. The thickness reaches about 670 µm in the periphery and is
slightly more than 500 µm in the central area. The cornea is composed of a stratified
squamous non-keratinized epithelium, limiting laminae Bowman membrane,
connective stroma tissue, Dua’s membrane, Descemet membrane and endothelial cell
monolayer.
Figure 2.2. The corneal structure. Dua’s membrane is located between the stroma and
Descemet layers.
Literature Review 13
Epithelium
The corneal epithelium is composed of approximately five to seven stratified
squamous cell layers in the centre and eight to ten in the periphery. The total thickness
is about 53.4 ± 4.6 µm (Reinstein et al., 2008). The epithelium conserves transparency
and refractive characteristics of an avascular connective tissue. This stratified layer
maintains metabolic activity, has a strong resistance to abrasion and has rapid healing
ability (Gatlin et al., 2003).
One characteristic of the epithelium is highly developed junctions with joined cell
membranes, which give great stability. Some processes called microvilli are located
on the outside surface of cells. These processes are between 0.5 and 1.2 µm in length
with an approximate thickness between 0.15 and 0.5 µm and are renovated in the cell
cycle (Gipson, 1994).
The corneal epithelium is a complex intercellular network that allows the eye to
withstand the pressures of abrasion when blinking. A diversity of molecules is
responsible for intercellular adhesion to the corneal epithelium. There are two
particular groups; one is cadherins, made of calcium-dependent glycoproteins, and the
second is integrins, proteins embedded in the cell membrane. In addition, the basal
cells are joined by microstructures that have the role of communication barriers and
there are essentially three different types: first, desmosomes are structures which
remain attached to neighbouring cells; second, tight junctions form an enclosed
system preventing the passage of molecules; third, gap junctions form ion channels
and hydrophilic molecules (Gipson, 1992).
Corneal epithelial cells bind to the basement membrane, highlighting the presence of
keratin filaments in the central zone of the cytoplasm which constitute the
hemidesmosomes. These are fixed anchoring fibrils located in the basement
membrane, composed of collagen and penetrating the stroma. At this level is a whole
network of microstructures which hold together the epithelium, which is subject to
multiple stresses (Kurpakus et al., 1992).
The central epithelium is free of melanocytes and mature antigen-presenting dendritic
Literature Review 14
cells (Langerhans cells). However, a combination of both mature and immature
dendritic cells is found in the peripheral epithelium (Cruzat et al., 2011).
All this extracellular structure helps maintain a barrier, allows passage of fluids from
the tear film to the stroma and protects the cornea from infections. The microvilli on
the surface of the outer cells are associated with the glycocalyx that adheres to the
mucin layer of the tear film and the ocular surface (Gipson and Argüeso, 2003).
In CLW, morphological endothelium cell changes such polymegethism and
polymorphism have been reported (Lemp and Gold, 1986).
Bowman's membrane
The acellular Bowman’s membrane is located immediately posterior to the corneal
epithelium. It is 8 to 17.7 µm thick and composed of collagen fibrils about 35 nm in
diameter (Tao et al., 2011). Their union with the stroma is undetectable, but there is a
clear delineation from the basement membrane. The great importance of Bowman's
membrane is its maintenance of corneal and optical stability (Jacobsen et al., 1984).
Stroma
The stromal layer represents 90% of the total corneal thickness. In the central part, the
thickness is about 500 µm and up to 700 µm in the periphery. The stromal structure is
mainly composed of collagen fibres. These are arranged in between 200 to 250
parallel layers, also called lamellae. All the fibres have an equal direction within each
layer, but between lamellas the orientation is oblique. The narrow diameter of the
fibres (30 to 38 nm) is a feature that contributes to transparency and is owed to the
size of the collagen molecule (Maurice, 1970).
The inter-fibre space contains proteoglycan-type keratin sulphate and dermatan
sulphate. The first is more evident in the central and anterior stroma and its role in this
space is to retain cations and water. The collagen molecules create binding bridges to
compensate for the separating force of the fibres with some elasticity to fit tensional
forces. This also explains the need to exert a force to separate stromal lamellae. The
arrangement of the fibres ensures uniformity throughout the structure of the corneal
dynamic. Proteoglycans bind to the fibres in an orderly collagenase (a proteoglycan
Literature Review 15
binds to a specific point of attachment), which is essential for sorting fibrillary
spacing (Scott, 1988).
The keratocytes, stromal cell components, are arranged between the plates and
maintain the structure. These cells are responsible for synthesizing collagen and
proteoglycans. The enzymatic characteristics of these cells allow the synthesis of
material and migration to the site of conflict to restore damaged structures.
In oedema conditions, the shape of the stromal changes, elongating the fibres and
changing the curvature of the anterior surface. In CLW, loss of keratocytes (Efron,
2007; Efron et al., 2002; Patel et al., 2001), opacities (Brooks et al., 1986; Efron,
2007; Pimenides et al., 1996), infiltrates (Carnt et al., 2009; Chalmers et al., 2012b;
Hickson and Papas, 1997; Holden et al., 1999) and neovascularization (Efron, 2012)
associated with oedema of the stroma have been reported.
Dua’s membrane
Dua’s membrane was detected in 2013 by Harminder Singh Dua during transplant-
related research using air bubbles. This layer is approximately 6 to 15 µm thick,
located between the stroma and Descemet’s layers. Although Dua’s membrane is very
thin, this membrane is strong enough to withstand approximately 700 mm of Hg of
pressure (Dua et al., 2013). Research with regards this membrane still ongoing to
confirm its importance in the corneal structure.
Descemet's membrane
Descemet's membrane is a thin homogeneous (6 to 11 µm) layer that stays attached to
the stroma. It is rich in the glycoproteins laminae and collagen that provide great
elasticity and friction. The resistance to traumatic or inflammatory injury of this layer
is greater than that of the stroma (Johnson et al., 1982). When there is an injury to the
epithelium from trauma or internal disease, the endothelial cells secrete collagen
fibres, creating a posterior banded layer on the Descemet membrane (Waring, 1982).
Overwear of rigid CL can increase a number of keratocytes in the pre-Descemet’s
layer (Curran et al., 1974).
Literature Review 16
Endothelium
The corneal endothelium is a monolayer of cells forming a cubic, hexagonal mosaic.
The ultrastructure shows adhesion to the specialized Descemet’s membrane that is
separate from the stroma. At its apical portion, the endothelium is in contact with the
aqueous humour that provides a smooth surface for optical quality conditions
(Bourne, 1983).
The nucleus of endothelial cells is large, and the mitochondrial material is abundant in
the cytoplasm. In smaller proportion, there is a rough and smooth endoplasmic
reticulum, ribosomes and Golgi. There is accumulation of ATPase in the cell borders
to control the energy transmitted to hydrate the stroma (Dawson and Edelhauser,
2010).
Unlike the corneal epithelium, the endothelial membrane lacks the ability of cell
renewal. This causes a loss of the cell population and thickness with age. Corneal
endothelial studies have been of interest to researchers since it is possible to evaluate
cellular structures and density. In young adults the cellular density is about 3,000
cells/mm², estimated to decrease by 500 cells/mm² in late adulthood, resulting in
around 2500 cells/mm² (Bourne, 2003). Endothelial loss is also associated with
polymegethism (diversity in size between cells), pleomorphism (diverse forms) and
increased polytonality associated with increased permeability (Sheng and Bullimore,
2007).
Morphological changes to the endothelium such endothelial blebs (Holden et al.,
1985; Inagaki et al., 2003; Vannas et al., 1984; Zantos and Holden, 1977),
polymegethism (Esgin and Erda, 2002; Hollingsworth and Efron, 2004; Nieuwendaal
et al., 1994; Wiffen et al., 2000), permeability (Chang et al., 2000; Dutt et al., 1989)
and cell loss (Dada et al., 1989; MacRae et al., 1994; McMahon et al., 1996; Setälä et
al., 1998) can be observed in CLW.
The main considerations in terms of corneal physiology are the barrier functions,
metabolism and pumping from the epithelium throughout the endothelium. A
disruption in any of these layers would lead to oedema with a loss of transparency.
Literature Review 17
2.1.3 The Conjunctiva
Structural layers
The conjunctiva is a mucosal membrane that covers the posterior side of the eyelids
and extends to the limbus onto the eye surface. Histologically, it is composed of a
layer of keratinized stratified epithelium and the substantia propria. The epithelium
has a variable number of cell layers, between three and seven. The apical cells interact
with the external environment via particle and bacteria phagocytosis and the secretion
of substances. The substantia propria consists of a highly vascularized connective
tissue with the presence of fibroblasts, lymphocytes, mast cells, plasma cells and
neutrophils (Brücke, 1847).
The two main functions of the conjunctiva are to provide mucins to the mucus layer
of the tear film and participate in the defence system of the ocular surface. The
conjunctiva is differentiated according to three distinct regions: palpebral, fornix and
bulbar conjunctiva.
The palpebral or tarsal conjunctiva adheres firmly to the back of the tarsus. It binds to
the skin at the edge of the eyelid and contains the crypts of Henle located in the upper
third of the inferior tarsal conjunctiva and the lower third of the upper tarsal
conjunctiva (Henle, 1860).
The conjunctival fornix or cul-de-sac is the fold that forms the conjunctival mucosa
passing the eyelids through the eyeball. Its projection on the eyelids shows a circular
shape, reaching above and below the upper and lower orbital rims and lateral to both
canthi. At the top of the fornix in the proximity of the bulbar conjunctiva are located
the Krause and Manz glands and the GCs (Metz, 1868).
The bulbar conjunctiva is thinner than the cornea and covers the exposed part of the
eyeball. It can be divided into scleral and limbal portions. The scleral portion is the
inner angle that forms the lacrimal caruncle and semilunar fold. The limbal portion
adheres to the corneal membrane to form the limbus. As a mucous membrane, the
conjunctiva is also composed of epithelial and stromal layers.
Literature Review 18
Figure 2.3. The conjunctiva structure.
Epithelium
The epithelial layer is formed by two to eight cell layers. In the tarsal conjunctiva and
the anterior fornix are about two cell layers, whereas the bulbar conjunctiva has
between six and eight layers. The basal cells are cuboidal and flattened or disrupted
by GCs (Krause, 1854).
Stroma or substantia propria
The deeper stromal layer consists of highly vascular connective tissue that is
separated from the epithelium by a basement membrane. Some authors subdivide the
stroma into two layers: one lymphoid layer which contains lymphocytes and elastic
fibres; and a deep layer six times thicker which leads to the tarsal plates and contains
vessels, glands and nerves (Metz, 1868).
Glands of the conjunctiva
The glands of the conjunctiva are classified into accessory lacrimal glands and mucin-
producing glands.
Literature Review 19
Accessory lacrimal glands
The glands of Krause are located at the superior fornix and are structurally similar to
the lacrimal glands (Krause, 1854).
Wolfring glands are also called glands of Ciaccio or tarsoconjunctival acinar glands.
These glands contributed to producing the aqueous portion of tear film (Ciaccio,
1873).
Mucin-producing glands
Henle crypts are present in the tarsal conjunctiva, close to the fornix. These glands are
folds or invaginations of the conjunctival epithelium (Henle, 1871), and their
glandular nature is under discussion. However, there are assumptions about their
immunological function.
The GCs are unicellular glands with numerous functions and characteristics that will
be explained in the following section.
2.2 CONJUNCTIVAL GOBLET CELLS
GCs are globular in morphology and loaded with dense mucin granules that contain
mucin-type MUC5AC. These cells have a wider apical part where the secretory
vesicles are accumulated; and a narrow basal part where the nucleus and organelles
are located. GCs are scattered among the epithelial lining of the organs, such as the
intestinal and respiratory tracts, the trachea, bronchi and bronchioles, larger
respiratory tract, small intestine, colon, and the conjunctiva. Conjunctival GCs have
been studied extensively with regard to mucin secretion and cell proliferation in
animal model ex vivo studies, but not specifically in the human conjunctiva because
of the difficulty of following a given cell after stimulation in vivo (Hodges and Dartt,
2010).
Literature Review 20
Figure 2.4. Goblet cell structure.
2.2.1 Development
In the human conjunctiva, GCs are assumed to develop from basal stem cells
(Kessing, 1968). The differentiation emerges from the fornix to the palpebral and
bulbar region between the 8th
and 9th
week of gestation (Miyashita et al., 1992).
Secretory granules and mitochondrial portions can be observed by the eleventh to
twelfth week (Sellheyer and Spitznas, 1988). During weeks 9 to 11, a spherical
structure with double layer membranes called autophagosomes can be observed. The
presence of these double layer membranes strongly indicates intracellular degradation
of cytoplasmic contents (De Duve and Wattiaux, 1966). It is assumed that some of
these cells are programmed to die during embryogenesis (Sellheyer and Spitznas,
1988).
2.2.2 Function
The main function of conjunctival GCs is to synthesize, store and secrete mucin
granules containing gel-forming mucin-type MUC5AC. These secretory granules also
contain glycoproteins including peroxidase (Iwata et al., 1976), trefoil peptides (TFF1
and TFF3) (Langer et al., 1999) and defensins (Haynes et al., 1999; McNamara et al.,
Literature Review 21
1999). Some GCs of the human conjunctiva have been reported to express small-
sized-molecule mucin-type MUC4 message of Ribonucleic acid (mRNA) as evaluated
under high resolution fluorescence (Gipson and Inatomi, 1998). Argüeso and Gipson
(2001) proposed a model of MUC5AC mucin packaging and secretion showing that
the molecule of mucin that is about to be released unfolds during secretion. Then the
MUC5AC is released onto the ocular surface, the molecules spread and associate with
the mucus gel along with other mucins expressed by cells of the squamous epithelium.
In order to balance mucin release by GCs into the tear film it is important to maintain
a number of cells in the ocular surface. The amount of mucin synthesized and stored
depends on the rate of mucin secretion and degradation (Hodges and Dartt, 2010).
2.2.3 Proliferation
EGF is known to play an important role in conjunctival GC proliferation as
demonstrated in culture studies on the rat (Gipson et al., 2003). In the presence of the
inflammatory cascade, the EGF receptor induces GC differentiation. Animal model
investigations have shown that after 1 minute of stimulation with EGF epithelial cells
begin to proliferate; after 18 hours, a second peak of proliferation GCs appears (Knop
and Knop, 2010). The real mechanisms stimulating human conjunctival GC
proliferation in a natural environment are unknown. However, once the stimulus of
these cells can be identified, the knowledge could potentially be used in a number of
treatments including improvement of CL comfort.
2.2.4 Secretion
GC secretion is released through the plasma membrane producing membrane-bound
vesicles in the apical portion of the cell. The body loses part of its cytoplasm in the
secretions without risk of exocytosis. Thus, the quantity of mucin secreted onto the
ocular surface depends on the GCD and the strength of the stimulus (Hodges and
Dartt, 2010). Measurements of conjunctival GC secretion have been reported in both
in vivo and ex vivo studies.
In vivo studies by Dartt et al. (1995) revealed that conjunctival GC secretion respond
to corneal and conjunctival nerves stimulation. These findings were ratified a year
later in an in vitro experiment using parasympathetic neurotransmitters acetylcholine
Literature Review 22
and vasoactine intestinal peptide pathways. Sympathetic transmitters did no
stimulated secretion in the in vitro study (Dartt et al., 1996).
Growth factors have also demonstrated response to conjunctival GC secretion. These
factors include EGF (Watanabe et al., 1993), vascular endothelial growth factor
(Joussen et al., 2003) and neurotrophic factors including factors 3 and 4 (Ghinelli et
al., 2003). Growth factors are secreted in the tear film by the lacrimal gland and
interact with GCs through the permeable membrane of the basal cells (Diebold et al.,
2001). All these experiments used specific receptors in order to stimulate the GCs of
rats, mice and humans in specific environments and conditions in order to obtain the
response of these cells. However, the stimulus of GC secretion in the natural
environment of the human conjunctiva is hard to explore and remains unknown.
2.2.5 Distribution
Early studies looking at the distribution of GCs in healthy human conjunctival
samples were carried out by Kessing (1968) using flat-mount preparations under light
microscopy. The distribution of GCD was reported in four quadrants (lower and upper
nasal; lower and upper temporal) including the fornix, bulbar and tarsal conjunctiva.
This author concluded that the higher GCD was in the two nasal quadrants and
approximately GCD ranges of 300 to 800 cells/mm² were observed in adults between
20 and 80 years old.
After Kessing, the literature focuses mainly on observations of morphological
changes in cells from the ocular surface in disorders and the development of grading
systems associated with these changes, called squamous metaplasia. Nelson and
Wright (1984), who were also studying morphological changes of the epithelial cells
developed a grading system that incorporated GCD values of the normal conjunctiva
over the bulbar and palpebral regions; these values were around 443 ± 266 and 1972 ±
862 cells/mm², respectively. The values were averaged from samples taken from the
interpalpebral and inferior palpebral conjunctiva in different individuals.
Other investigators have used different regions of the conjunctiva to report GCD.
Some reports in the literature have averaged samples from different parts of the
conjunctiva (palpebral, interpalpebral and bulbar) and different regions such as upper,
lower, temporal and nasal.
Literature Review 23
For the purpose of the present review of the literature and in view of the reports of GC
distribution in healthy participants, a total of 61 reports on GCD values were grouped
into three data sets as shown in Figure 2.5. The first dataset (blue bars) included
values of GCD from the upper bulbar conjunctiva including upper-temporal
(Ciancaglini et al., 2008; Karalezli et al., 2011; Karalezli et al., 2009; Mrugacz et al.,
2008; Rivas et al., 1991), upper-central (Adar et al., 1997; Aksünger et al., 1997;
Çakmak et al., 2003; Murube and Rivas, 2003; Paschides et al., 1991; Rivas et al.,
1991; Rivas et al., 1993; Rodriguez-Prats et al., 2007; Rodriguez et al., 2007; Rojas et
al., 1993) and upper-nasal (Rivas et al., 1991).
The second dataset presented (red bars) reflects the medial bulbar conjunctiva
including reports using averages of nasal and temporal (Matsumoto et al., 2008;
Murube and Rivas, 2003; Nelson and Wright, 1984; Rivas et al., 1991; Satici et al.,
2003; Wang et al., 2007) as well as single values (Bai et al., 2010; Kim et al., 2007;
Rivas et al., 1991; Rojas et al., 1993; Rummenie et al., 2008; Solomon et al., 2004).
The last set (green bars) were reports of the lower bulbar conjunctiva (lower-nasal)
(Dogru et al., 2003; Dogru et al., 2001; Dogru et al., 2000; Dogru et al., 2002; Rivas
et al., 1991; Yoon et al., 2005; Yoon et al., 2004), lower central (Moreno et al., 2003;
Murube and Rivas, 2003; Nishida et al., 1995; Rivas et al., 1995; Rivas et al., 1991)
and lower-temporal (Filippello et al., 1997; Rivas et al., 1991; Rodriguez-Prats et al.,
2007; Rodriguez et al., 2007; Tseng et al., 2001).
All these reports used different techniques of GC assessments in healthy individuals.
No data from any of the three groups showed significant differences between the
mean values (p > 0.05). However, higher values were observed in the lower bulbar
conjunctiva 734 ± 621cells/mm² compared with the medial and upper bulbar (GCD
519 ± 458 and 511± 514 cells/mm², respectively). This analysis also reveals that
approximately 82% of the mean values from these reports on the bulbar conjunctiva
were GCD values under 550 cells/mm².
Literature Review 24
Figure 2.5. Histogram shows the distribution of GCD values reported in the healthy bulbar conjunctiva. The GCD values under the line represent
82% of 61 values reported. The blue, red and green bars represent GCD of nasal and temporal, upper and lower bulbar conjunctiva respectively.
0
200
400
600
800
1000
1200
1400
1600
1800
2000
GC
D c
ell
s/m
m²
Av nasal and temporal Temporal Nasal Upper-temporal Upper-central
Upper-nasal Lower-temporal Lower-central Lower-nasal 82% values under 550 cells/mm²
Literature Review 25
2.3 GOBLET CELL ASSESSMENT TECHNIQUES
This section will explain briefly how GCs have been collected since 1968. However,
it concentrates on the techniques that were used in the present study and the
association between the measurement of GCD with CIC and LSCM.
2.3.1 Biopsy
Histology studies of the GCs from conjunctival tissue were used widely for over a
decade between late 1960s to the mid-1970s before the development of the CIC
technique. As mentioned before, Kessing was the forerunner of conjunctival
observation using cadaveric tissues of the conjunctiva prepared on flat mounts and
observed under a light microscope.
The biopsy technique allows the observation of tissue that is usually processed
(dehydrated, cleared and infiltrated), embedded and sectioned before staining and
prepared on slides to be examined by light or electron microscopy, typically in a
cross-sectional view. However, transparent tissues such as the cornea and the
conjunctiva can be observed under microscopes without staining.
Observations of GC using flat-mount biopsies of the conjunctiva under light
microscopy have been described previously by many authors (Greiner et al., 1981;
Gwynn et al., 1993; Kawano et al., 1984; Kessing, 1968; Vujković et al., 2001;
Yamabayashi and Tsukahara, 1987). The descriptions of GCs under light microscopy
indicate that these cells can be seen throughout the epithelial layer of the conjunctival
tissue as single units predominantly in the superficial layers (Doughty, 2012a). In
sections that are stained, GCs can look different according to the stain used. Positive
stains for mucus detection such as PAS show GCs as pink, round to oval in shape, but
neither the nucleus nor the cell body is distinctive. In contrast, alcohol-based stains
such as Giemsa stain allow observations of the nucleus of the cells and comparison of
the nucleus and the cytoplasm size at high magnification.
With electron microscopy GC samples are treated differently. Usually, they are fixed
in glutaraldehyde to observe internal mucin granules of the cell at a very high
magnification (2000X). They appear to have a depression in the apical portion that is
Literature Review 26
in contact with the ocular surface. However, this is difficult to see in GCs located
deeply in inner layers. GC shape visualized using electron microscope depends on the
quality of the tissue process before imaging (Doughty, 2012a).
2.3.2 Conjunctival Impression Cytology
This is a mildly invasive technique involving the removal of cells from the ocular
conjunctiva for examination under a microscope or isolation of cells for analysis with
flow cytometry (FC) and PCR.
The CIC technique is attributed to Egbert et al. (1977) who after touching the bulbar
conjunctiva with filter acetate, saw under a microscope some points that mucin had
formed on the filter paper. The filter was then stained with a chemical solution of PAS
and the locations of the mucus were taken to be an indicator of GCs. This observation
was consistent with the description of GCs with the biopsy technique. However, the
difference is that the observations under a light microscope using CIC provide a
frontal or coronal plane of view. Thus, GCs can be counted and reported as cells/mm²
Staining methods can vary in this technique according to the filter used to collect the
cells. Conventional cellulose acetate filters allow the observation of cells under a light
microscope using coloured stains. For immunofluorescence staining, however, the
filter must have specific properties such as mixed cellulose esters and larger pore size.
A few reports in the literature have mentioned that different filter types can improve
sample consistency and cell attachment (Albietz, 1999; Doughty, 2012b). However,
studies using conventional cellulose acetates applied pressure and increased the time
of application during the sample collection to obtain the same outcomes as mixed
cellulose esters.
Various grading system methods have been used to report GCD. In 1984 Nelson and
Wright proposed the now most commonly used grading scale that considers the
development of the epithelial cells and includes specific variations in GCD associated
with epithelial cell changes. This grading system is comprised of four grades: grade
zero is a strongly positive response involving GCD with more than 500 cells/mm²;
grade one is slightly positive involving 350 - 500 cells/mm²; grade two is a moderate
negative response involving 100 - 350 cells/mm²; and grade three is considered as a
Literature Review 27
negative response and represented GCD samples that has fewer than 100 cells/mm². A
year later Tseng (1985) suggested a specific grading system based on variations and
keratinization of epithelial and GCs. The grading system was later divided into six
stages which involved correlation of the number of GCs with the nucleus-cytoplasm
ratio of squamous cells. Connor et al. (1991), correlated histopathological changes
and based a grading system on the thickness and the extent of the metaplasia in the
conjunctival epithelium and GC estimates. Nuclear changes of GCs were
contemplated in the grading system described by (Aragona et al., 1998). Frequency,
morphology and GCD were considered in this scoring system for CIC.
New technology was used by Pisella et al. (2000), who conducted research using FC
after isolation of GCs by CIC techniques. This application used antibodies directed at
the human leukocyte antigen d-related (HLA-DR) and intercellular adhesion molecule
type 1 (ICAM-1) (CD 54) as inflammatory markers, and at the peptidic core of the
conjunctival mucin-type (MUC5AC) for mucus type and GC detection.
Another modern application used in association with CIC techniques is PCR. This
application allows the analysis of mRNA specific to a particular mucin type, in this
case MUC5AC for GC detection (Corrales et al., 2009).
2.3.3 Laser Scanning Confocal Microscopy
In the history of technological development in ocular diagnostic equipment, LSCM
has emerged as a valuable tool for evaluating cellular layers of tissue on the anterior
surface of the eye in vivo, thus allowing the analysis of GCs located in the
conjunctiva. Previous studies have reported observations of GCs assessed by LSCM
as highly reflective cells (Guthoff et al., 2006), approximately 30 µm in diameter
(Kobayashi et al., 2005), round to oval in shape and sometimes visible grouped along
the conjunctival epithelium (Messmer et al., 2006). However, there are some areas of
disagreement in the literature regarding the interpretation of LSCM images of the
bulbar conjunctiva. The cellular structures appear in black and white, and the
morphological description is based on comparing sizes, shapes and translucent or
opaque structures that can vary depending on the position and depth of the LSCM on
the conjunctiva (Efron et al., 2009).
Literature Review 28
GCD calculation using LSCM is a faster technique than using any other method for
GC assessment due to its cell count mode using Heidelberg software. Another
advantage of using this method is the ability to focus in a Z plane, which gives the
choice of not only to explore different depths of tissue but also to compare and
correlate the technique accuracy with the conventional way of GC assessment CIC.
2.3.4 Correlation between CIC and LSCM
Both impression cytology using Giemsa stain and LSCM are methods for cell
morphology evaluation of the ocular surface. The grading scale system developed by
Nelson and co-workers reflects metaplasic changes to epithelial cells as well as
changes in the number of GCs using CIC. This scale has been used to identify cells on
the ocular surface using CIC and LSCM techniques in eyes treated with both
preserved and preservative-free glaucoma therapies. A positive correlation of GCD
using LSCM and CIC has also been demonstrated in people with Sjögren syndrome (ρ
= 0.908; p = <0.05) (Hong et al., 2010) and chemical burns (ρ = 0.946; p = 0.000) (Le
et al., 2010). In addition, correlation analysis has also been done in participants with
keratoconjunctivitis measuring inflammatory cells (R = 0.97; p = <0.05) (Wakamatsu
et al., 2009).
Positive correlations between the two techniques with regard to GCD have never been
reported in healthy participants. The CIC technique has been widely used for the past
three decades to report GCD. However, limitations of this technique have been
observed by many authors therefore it is important to establish and understand the
degree of association between the CIC and less invasive techniques in order to
facilitate future GC assessment.
Changes of GCD occur in response to CL wear. The following sections explore the
manifestations of the GCD in response to CLW described in the literature.
2.4 CONTACT LENSES
2.4.1 History
CLs are medical devices mainly used for visual improvement. Cosmetic, therapeutic
and functional reasons are other factors that can motivate people to wear CLs instead
of spectacles. Data from 2004 showed that approximately 125 million people (2%) of
Literature Review 29
the worldwide population wear CLs (Barr, 2005). The 2014 Annual Report on CLs
showed an increase of approximately 1.2 million new CL wearers in the United States
alone from 2004 to 2014 (Nichols, 2015).
The development and evolution of CLs began with rigid lenses; then rigid gas
permeable (RGP) lenses appeared, followed by soft hydrophilic CLs with a variable
percentage of water, and finally the latest silicone hydrogel lenses that complement
the advantages of rigid materials with hydrophilic siloxane.
Rigid CLs were developed between the 1930s and late 1970s and were originally only
composed of polymethylmethacrylate (PMMA). These lenses had excellent optical
properties and were easy to clean and required minimal care. However, PMMA was
not good in terms of oxygen permeability, which reduced the tolerability of the lens.
Therefore, in the late 1970s silicone or fluorine was added to the PMMA to increase
the oxygen permeability and so more flexible RGP lenses with an adequate level of
oxygen transmissibility were made possible.
Soft lenses began to be manufactured after the decade of 60s. The main component of
most of these lenses is hydroxyethyl-methacrylate (HEMA). Water percentage ranges
mostly between 38% and 85%. The higher the proportion of water, the better its
oxygen permeability. The concentration of water can be increased by attaching
HEMA to other materials.
Soft CL can be classified into two groups according the lens material: conventional
and silicone hydrogels.
In conventional hydrogels, the oxygen is transported through the aqueous phase of the
lens to the cornea and permeability to the oxygen is directly proportional to the water
content of the material. The best oxygen transmissibility in conventional hydrogels is
CLs with high water content along with minimal thickness. However, thin lenses of
high water content are not well tolerated, and lead to cause a superficial punctate
keratopathy (Mobilia and Foster, 1978). This phenomenon is caused by the rapid
dehydration of lenses which consequently adhere to the epithelial cells causing
corneal surface disruption. When the epithelium disrupts, it facilitates bacterial
invasion of the cornea.
Literature Review 30
This deficiency motivated the industry to investigate the possibility of creating new
hydrogels with higher oxygen transmissibility while maintaining the characteristics of
conventional hydrogel lenses. This has led to the invention of a new generation of
silicone hydrogels. These hydrogels are made with new polymers that not only absorb
water but are also permeable to oxygen.
Silicone hydrogels are relatively new materials that facilitate the manufacture of soft
hydrophilic lenses with high oxygen permeability. Whereas the oxygen permeability
of conventional hydrogels depends essentially on the water content, in silicon
hydrogels oxygen permeability depends on the chemical structure of its silicone
phase.
These new hydrogels have produced CLs with approximately 50% water content, and
transmit sufficient oxygen to satisfy the corneal physiology with the closed eye. These
lenses can be used for overnight wear.
In summary, it can be concluded that CL materials evolved over the years to promote
ocular surface health and CL comfort. However, symptoms of dryness in CL wearers
are widespread as demonstrated in three reports from different countries (Doughty et
al., 1997; Lowther, 1997; Orsborn and Robboy, 1989). In the following section, DE
related to CLW is discussed.
2.4.2 Contact Lens-Induced Dry Eye Symptoms
Extensive evidence in the literature has determined that approximately 50% of the
contact lens wearing population withdraws from CL wear most frequently due to
symptoms of dryness. According to the recent report of the 2013 TFOS Contact Lens
Discomfort workshop, the term “CL-induced DE” should be used to describe the
pathophysiology of those without pre-existing DE condition (Nichols et al., 2013).
In other words, symptoms of DE are not present in the absence of CLs indicating that
CLW can induce subclinical DE causing symptoms of dryness only when lenses are
in the eye.
Alteration of the tear film is especially problematic during CLW because of the need
to maintain the optical surface of the lens as well as ocular surface stability. In soft
Literature Review 31
CLW, constant wetting is essential to maintain the elasticity and transparency of the
CL. There is evidence that the development of DE symptoms during CLW could
potentially be due to changes in structure and lipid production of the meibomian
glands (Farris, 1986), leading to a thinner lipid compound and affecting the
osmolarity of the tear film causing evaporation of tears (Villani et al., 2011b) and
dehydration of the lenses (Nichols and Sinnott, 2006). Potential mechanisms of CL–
induced DE include tear film evaporation (Guillon and Maissa, 2008), inflammation
(Pisella et al., 2001), an increase in the osmolarity (Gilbard et al., 1986) and
dehydration of the lens (Arita et al., 2009). However, the etiology of CL–induced DE
is still unknown and further studies need to be conducted.
2.5 DRY EYE
Dry eye is a condition that affects almost half of the CL population worldwide. When
conducting research trials using CLs it is important to determine any sign and
symptoms of DE in order to avoid bias and to deliver reliable results.
2.5.1 Dry Eye Diagnosis, Criteria and Definition
The term ‘dry eye’ is difficult to define because, regardless of the numerous causes,
the associated clinical manifestations vary greatly in intensity even over time in the
same patient. DE symptomatology may not correspond with the signs observed by the
practitioner. Subjective symptoms combined with assessment of objective evidence
forms the basis of diagnostic parameters. Because of a frequent lack of correlation
between signs and symptoms, marginal cases of DE may go unnoticed and
undiagnosed (Nelson, 1995).
Diagnosis of DE is important to be considered as one of the exclusion criteria when
fitting lenses in clinical trials in order to avoid confounding factors related to signs
and symptoms of CL-induced DE. The assessment of DE includes the following: (1)
symptoms, (2) tear instability, (3) reduced integrity of the ocular surface, and (4)
reduced tear volume as recommended by DEWS 2007.
The DEWS 2007 defined DE as ‘a multifactorial disease of the tears and ocular
surface that results in symptoms of discomfort, visual disturbance, and tear film
instability with potential damage to the ocular surface. It is accompanied by increased
Literature Review 32
osmolarity of the tear film and inflammation of the ocular surface’. This study uses
the DEWS definition and criteria for DE diagnosis and evaluation.
2.5.2 Dry Eye Evaluation
The following recommendations by the DEWS dictate the sequence of tests to be used
in DE assessment, according to category. The DE evaluation should be performed in
the sequence that best preserves the integrity of the tests.
Symptom questionnaires
A number of questionnaires have been developed to capture DE symptoms, to explore
the epidemiology of the condition (Doughty et al., 1997), to diagnose sufferers
(McMonnies and Ho, 1987) and to assess treatment effects (Schiffman et al., 2000).
The time taken to administer a questionnaire influences its selection for general
clinical use and research trials. Self-reported symptoms of DE are valuable in view of
the lack of correlation available with an objective test. Validated questionnaires allow
rapid assessment and ensure consistency in the collection of relevant information. The
responses to the questions have assigned values, allowing the severity of the DE to be
rated and the effectiveness of treatments to be monitored.
Two statistically validated questionnaires have emerged in the field of the evaluation
of DE in both CL wearers and non-CL wearers that were also suited for this study.
The 5-Item Dry Eye Questionnaire (DEQ-5) was developed for epidemiologic and
clinical studies to measure symptoms of ocular irritation in patients with aqueous tear
film deficiency (Chalmers et al., 2010). It consists of 5 items designed to measure
prevalence, frequency, diurnal effects and severity and intrusiveness of symptoms
(Report of the International Dry Eye WorkShop (DEWS), 2007). The Contact Lens
Dry Eye Questionnaire-8 (CLDEQ-8) was designed by the same group of researchers
with the aim of developing a screening questionnaire for soft CL-induced DE. It is
based on 8 items that measure not only symptoms but also changes in the condition
after treatment (Chalmers et al., 2012a).
Non-invasive tests of tear stability
Non-invasive testing stability is performed without touching the tear film and the
ocular surface. These tests are more valid than traditional tests because ocular dyes
Literature Review 33
have the potential to destabilize the tear film and reduce the measured value (Patel et
al., 1985). A non-invasive test that reflects the view of the tear film is the tear break-
up time (TBUT) test. In this test, the mires of various ophthalmic instruments such as
the keratometer can be used. The time between a blink and the first sign of alteration
or interruption of view while the patient refrains from blinking is the time of tear
thinning. Tear film stability depends on many factors, such as blink reflex
mechanisms, the presence of healthy lacrimal glandular tissue, and a structurally
intact tear film. All three layers of the tear film contribute to tear stability. Disorders
of any of the tear layers can strongly affect tear film stability. Non-invasive methods
include tear film lipid layer interferometry and the tear evaporation test. In this study
non-invasive tear break-up time; (NIBUT) is used to assess tear film stability non-
invasively.
Measurements of tear film stability are inherently variable. Therefore, an average of at
least 3 values of each eye should be recorded on average. In general, non-invasive
stability values are higher than those measured with dyes such as fluorescein.
Invasive tests of tear stability
Assessment of tear film stability can be achieved by fluorescein visualization of the
rupture of the tear film. The amount of sodium fluorescein instilled should be
minimal, ideally about 1 μl, using cobalt blue light plus a Boston filter (yellow). The
fluorescence visualization is greatly improved by these filters. The concept of TBUT
was first introduced by Norn (1969), who instilled sodium fluorescein using a
moistened strip or a pipette and observed the tear film with a biomicroscope, cobalt
blue light, and a Wratten 12 yellow barrier file (Cho and Douthwaite, 1995). The
subject was told to avoid blinking, and TBUT was defined as the time interval
between a complete blink and the appearance of the first observed break,
discontinuity, or dry spot in the tear film following a blink. Break-up occurs most
frequently in the inferior or central cornea (Elliott et al., 1998). In healthy eyes, TBUT
values range from 3 to 132 s, with an average of 27 s (Norn, 1969). Values <10 s
suggest an abnormal tear film (Mengher et al., 1985); values of 5 to 10 s are
considered as marginally abnormal, and values <5 s are associated with DE symptoms
(Pflugfelder et al., 1998).
Literature Review 34
Tear film osmolarity
The measurement of tear film osmolarity is claimed to be an important and relevant
measure of both DE and CL-induced DE. This test is carried out using a tear
osmometer that uses the resistance of an electric circuit in combination with
sophisticate calculations to acquire tear film osmolarity. A small tear sample is
obtained using a pipet that is then automatically transferred into the circuit surface.
The result is obtained in seconds after the transfer (Sullivan et al., 2012). A number of
studies suggest that the diagnostic cut-off of this test is ≥ 316 MOsm/L (Baudouin et
al., 2013; Sullivan et al., 2012; Tomlinson et al., 2006).
Ocular surface damage
Staining of the ocular surface is a convenient technique for evaluating the integrity of
the corneal and conjunctival epithelia that can be used to identify compromised
epithelial cells (Korb, 2000). Irregular staining of the ocular surface is recognized as
one of the most common DE signs and is also associated with interpalpebral surface
damage, tear instability, and tear hyperosmolarity. However, not all CL wearers with
ocular surface damage develop symptoms of CL discomfort. The severity of such
surface damage can be quantified using vital dyes such as fluorescein, rose Bengal
and Lissamine green (Korb et al., 2008)
Hyperaemia of the palpebral and bulbar conjunctiva can be observed and classified
according to a standardized scale. There are several clinically acceptable scales for the
evaluation of the ocular surface. However, the most commonly used grading scales
are the Efron Grading Scales for Contact Lens Complications and the Cornea and
Contact Lens Research Unit (CCLRU) (Efron et al., 2001). These scales quantify the
degree of redness of the bulbar conjunctiva in any inflammatory condition of the eye.
Measurements of tear volume
Tear volume was first measured by Schirmer (1903), who used filter paper to collect
tear secretions. Strips of special filter paper (35 mm × 5 mm) are placed in the lower
lid. Anaesthesia is optional. The commonly acceptable definition for DE is less than 5
mm of wetting in 5 min whereas 5 to 10 mm is described as borderline DE; normal
wetness is more than 10 mm of wetting.
Literature Review 35
The phenol red thread (PRT) test was refined by Hamano and Bode (1985), who used
a thread impregnated with phenol red, which is pH-sensitive and changes from yellow
to red over the pH range of normal tears. This change in colour helps to identify the
length of cotton thread wetted by tears. PRT still requires the thread to be hooked over
the lower lid but can be completed in 15 s (Tomlinson et al., 2001). The PRT test is
said to provide an index of tear volume, which is related to the tear secretory rate and
can therefore be used to diagnose aqueous-deficient DE disease (Bron, 2001).
Tear interferometry
Tear interferometry is a modern technique that has been used as a non-invasive
method to evaluate the tear lipid layer (Korb, 2002) and diagnose aqueous tear
deficient DE (Goto and Tseng, 2003). This test examines the superficial tear lipid
layer through a tear interference camera. Interference images are used to grade DE
severity according to the thickness of the lipid layer (King-Smith et al., 1999). This
test has also been performed in soft CL wearers where high water content
conventional hydrogel lens wearers were found to have reduced lipid interference and
an unstable aqueous layer compared with low water content CL wearers (Maruyama
et al., 2004).
In summary, numerous methods exist for assessing the tear film. In some cases,
however, the test procedure may influence the parameter under investigation by
inducing reflex tearing. The aim in recent years has been to develop and promote the
use of research methods in the least invasive tear film examination. Thus, the
evaluation of the tear film should be assessed in its ‘physiological’ state as possible
(non-invasively). For all these reasons, to explore the effects of CL-induced DE
symptoms in this study, DE diagnostic tests were performed at the baseline for DE
exclusion criteria. The diagnosis of DE at baseline was based on a combination of
subjective and objective tests. Accordingly, this study included the administration of
questionnaires for both contact and non-contact-lens wearers, the assessment of tear
film volume and tear stability, and the evaluation of ocular surface integrity, in
accordance with the criteria for DE diagnosis suggested by DEWS 2007. For the
follow-up visits symptomatic and asymptomatic participants of CL group were sub-
grouped according to the soft CL wearer questionnaire CLDEQ-8.
Literature Review 36
2.6 FACTORS AFFECTING GOBLET CELL DENSITY
2.6.1 Factors influencing goblet cell differentiation
The Notch and Wnt cascades of secreted signalling molecule families have been
shown to be involved in epithelial cells and GC differentiation of the mice (van Es et
al., 2005). Studies have demonstrated that the Notch signalling pathways have an
involvement with conjunctival GC differentiation. Zhang et al. (2015) demonstrated
that the inhibition of a protein that increases gene expression (coactivator) can induce
abnormal hyperplasia and desquamation of the conjunctival cells of the mice
including loss of GCs (Zhang et al., 2015). Another factor that plays a role in the
differentiation of GCs is known as the Wnt pathway cascade. Studies of Wnt pathway
responsive genes have shown a significant reduction of GCs in mice, suggesting that
the Wnt antagonist is regulating a Wnt pathway cascade involved in GC
differentiation in the conjunctiva (Gipson, 2016).
2.6.1 Ocular surface diseases affecting goblet cell density
Varied forms of DE disease show a reduction of conjunctival GCD. Many etiologies
including, keratoconjunctivitis sicca, blepharitis and cicatrizing diseases, such as
ocular cicatricial pemphigoid and Stevens Johnson syndrome demonstrated lower
GCD than healthy controls using CIC (Nelson and Wright, 1986).
Expression of MUC5AC has been demonstrated to be reduced in Sjögrens syndrome
DE (Argüeso et al., 2002) as well as atopic keratoconjunctivitis (Dogru et al., 2008).
In these two conditions the number of GCs has also been demonstrated to be reduced
(Hong et al., 2010) (Dogru et al., 1998).
Patients with graft versus host disease DE were also shown to have decreased GCD
compared to those with non-DE allogeneic hematopoietic stem cell transplantation
(Wang et al., 2010).
CIC technique has been used to assess the number of GCs and squamous status of
conjunctival samples in Avitaminosis A condition. This condition is also known to
affect the epithelial health of the ocular surface and was demonstrated to cause
reduction of GCs in the conjunctiva (Natadisastra et al., 1988).
Literature Review 37
In conclusion, many forms of ocular surface disease with presence of DE alter the
number of GCs. The reason why GCD is reduced in different etiologies of DE still
unclear and further studies need to be done with regards this matter.
2.6.2 The effect of contact lens wear on goblet cell densities
GCD has been reported to decrease as a result of CLW (Doughty and Naase, 2008;
Moon et al., 2006; Tomatir et al., 2008). Also changes inside the GC such as decrease
in the nucleus-to-cytoplasm ratio (N/C ratio) (Doughty, 2011b), increase in acidic
mucin, reduction in cell size (Tseng et al., 1984) and a decrease in mucin and protein
content in the tears of CL wearers (Yasueda et al., 2005), have been reported. The
main scope of this review is the impact of CLW on conjunctival GCD. The following
subsections ‘Studies demonstrating a decrease in GCD with CLW’ and ‘Studies
demonstrating an increase and no changes in GCD with CLW’ explain in detail the
changes reported in the literature regarding GCD of CL wearers using the CIC
technique. Only one report of GCD assessed by LSCM in a cross-sectional pilot study
was found in the literature – that of Efron et al. (2010a)
Studies Demonstrating a Decrease in GCD with CLW
In the mid-1980s, Götz et al. (1986) documented nuclear changes in conjunctival cells
and a decrement in the GCD of CL wearers compared with the non-CL wearing
control group. Using the CIC technique, the authors demonstrated differences in
wearers depending on the material of the CL (PMMA, RGP and soft lenses). Four
years later, Saini et al. (1990) concluded that GC loss and squamous metaplasia of the
conjunctival epithelium seemed to increase with the duration of CLW, demonstrated
by the comparison of participants with more than 1 year of CLW vs. short-term CLW
(less than 1 year of CLW). They also reported impression cytological changes in hard
CLW.
Knop and Brewitt (1992b), recruited 14 participants into CLW and observed
alterations of normal conjunctival epithelium in response to CLW in a longitudinal
fashion study. This study revealed not only changes in morphological epithelial cells
but also decrease in GCD compared with the non-CL wearing control group. GCD
values before CLW from upper and lower bulbar conjunctiva were 194 ± 120
cells/mm², respectively. After 6 months of CLW to the GCD has reduced to 133 ± 46
Literature Review 38
cells/mm², respectively. The researchers also reported that the majority of specimens
collected by CIC were monolayer and that the lower values of GCD compared with
those in other reports was attributed to poor cell attachment.
The CIC in the study of Knop was performed monthly, however; reports of GC values
are only given for baseline and final visit. In addition, Adar et al. (1997) assessed the
superficial cells of the conjunctival epithelium in 25 rigid and soft CL wearers, and
observed dramatic changes in superficial cells compared to the control group (non-CL
wearers). However, there were no significant differences in conjunctival GCD
between soft and rigid CL wearers (12.1 ± 1.7 cells/mm² vs 9.4 ± 1.5 cells/mm²,
respectively).
Adar et al. disagreed with the findings of Saini et al., claiming that there was no
correlation between average duration of CLW and GCD. Similarly, Aragona et al.
(1998) demonstrated both a significant reduction in GCD of symptomatic compared
with asymptomatic CL wearers and significant differences in GCD between soft and
rigid CL wearers. Albietz (2001), reported a significant reduction in GCD in CLW
and attributed this reduction to a mechanical influence on conjunctival squamous
metaplasia.
Anshu et al. (2001) modified the CIC technique by using filter dissolvers, leaving the
cells on the slide immediately after the sample was collected and completing the
staining procedure on the slide. These techniques ensured better cytological
evaluation and preservation of the cytological material.
Changes in the conjunctiva as a consequence of using silicone hydrogel lens were first
reported by Şengör et al. (2002), who demonstrated less GCD in long-term (7.7 ± 3.3
years) CL wearers vs. non-CL wearers. Finally, Simon et al. (2002), performed CIC in
28 participants fitted with RGP and soft CLW and the follow-up examinations
showed a significant decrease in CIC over a period of 6 months in both materials.
However, soft CL wearers showed more dramatic reduction.
The two longitudinal studies by Knop and Brewitt (1992b) and Simon et al. (2002)
showed dynamic changes in GCD over time (6-month to 1-year period) in CLW. Only
one study reported GCD using cells per unit area (mm²). In conclusion, both cross-
Literature Review 39
sectional and longitudinal studies on GCD of CL wearers demonstrated a decrement
in GCD regardless of CL material and replacement. More work should be done using
validated methods looking for differences of GCD in new CL materials and different
schedule replacement.
Studies Demonstrating an Increase and No Changes in GCD with CLW
Connor et al., 1994 performed a modified technique based on repeating the peeling
off procedure three times in the same sampling area using acetate filter paper to obtain
more cell attachment (Connor et al., 1994). These researchers reported a nearly 2-fold
increase in GCD in patients who had never worn CL before and who were fitted with
soft CL for a period of 6 months. They concluded that the change in GCD was an
adaptive response of the conjunctival surface to the daily use of CL. These reports
were given in percentage values of the total number of cells in the sample (4.19% at
baseline to 7.84% at 6-month visit) and there were statistically significant increases;
however, there was no control group and a reasonable power was not reached with the
population sample (N = 18).
Later, in 1997, the same group of investigators reported a statistically non-significant
increase in GCD in participants that were fitted with daily lens wear. This time they
reported a power measure of 0.94 for the data analysed. GC values were reported in
percentages and at baseline (3.23% ± 0.36% SEM); there were fluctuations over time
but they were never higher than baseline values. Interestingly, the lower values were
observed at the 6-month visit (2.57%) (Connor et al., 1997), which indicated a
decrease in GCD that was not statistically significant. Therefore, there was no
statistically validated change over 6-month period in this report.
Pisella et al. (2001) used FC to isolate GCs and other inflammatory cells collected
using the CIC technique. This study revealed no significant differences in GCD in soft
and rigid CL wearers compared with the control group. The reports of this study were
given in levels of fluorescence that showed a non-statistically significant lower value
of GCD in the two CL groups compared with the controls.
Lievens et al. (2003) used the same technique as Connor et al. (1994) and (1997),
whereby GCD was compared between wearers of silicone and conventional hydrogel
lenses. This study showed a higher value in GCD in both groups after 6 months of
Literature Review 40
CLW. However, there was a slight difference between the materials whereby silicone
hydrogels showed less increment in GCD. This group concluded that silicon
hydrogels are less irritating to the ocular surface than conventional hydrogels.
Hori et al. (2006) used a PCR application but surprisingly obtained no significant
changes in mucin production. This study did not evaluate GCD; however, these
authors related mucin production to the number of GCs.
Corrales et al. (2009) Interestingly, Corrales et al used the CIC technique in
conjunction with PCR to analyse differences of mucin-type MUC5AC between low-
and high-water content lenses of CL wearers. The results of this study showed an
increase in MUC5AC density after 1 year of use of high water-content hydrogels.
These reports were given in mRNA expressions (Log 0.8) indicating a significant
increase of MUC5AC after 1 year of CLW.
Efron et al. (2010a) conducted a pilot study where average GCD was compared
between 11 CL wearers and 11 non-CL wearers using LSCM. They observed no
statistically significant differences in GCD between the two groups, when four
cardinal points of the bulbar conjunctiva were averaged. However, the sample size
was small.
Table 2.1 summarizes the studies conducted to date where GCD is assessed in CLW.
The table provides a detailed differentiation between techniques and methods used by
previous studies about GCD assessed by CIC and LSCM techniques in CL wearers.
The varied results regarding the effect of CLW on GC are shown in Table 2.1. The
weight of evidence appears to be in favour of a reduction in GCD with CL wear.
Three studies Knop and Brewitt (1992a), Simon et al. (2002) and Tomatir et al.
(2008) showed decrease GCD after fitting participants into CLW for a period between
5 months to 1 year. On the other hand, cross-sectional studies of long-term (more than
13 months) CL wearers showed decrease GCD compared to healthy participants
(Adar et al., 1997; Doughty and Naase, 2008; Moon et al., 2006).
There are some variations in the methodology of the CIC technique that may relate to
the discrepancies in the results of GCD in CLW. The standardization of the technique
Literature Review 41
and the reports of all the studies analysed showed differences in terms of materials
used in the sample collection and staining procedure. There are also variations in the
CIC technique (FC and PCR), grading schemes and sampling consistency. There is a
lack of information in the majority of reports with regard to numbers of GCD and
units reported. Thus, the possible reasons for the inconsistency of GCD in CLW are
still unknown.
In summary, it is difficult to draw a conclusion from the current literature regarding
the impact of CL wear on GCD when there is such variability in terms of the
methodology and the units for GCD used in the literature. It could have many causes,
such as the quality of cell attachment to the acetate filter, type of filter used for the
CIC, region of the conjunctiva for sample collection, technique used for analysis, age,
gender, and even ethnicity.
Literature Review 42
Table 2.1. Goblet cell density assessed by conjunctival impression cytology in contact lens wear; ordered by direction of change.
Year Author N CL Type Method CL Wear
(years) Location Grading Method
Changes in
GCD
GCD
(cells/mm²) Other Findings
1986 Götz et al. 25 PMMA, RGP, SOFT CIC n.s Upper bulbar ↑ Squamous
metaplasia + GCD Decreased n.s
Snakelike chromatin present in keratoconjunctivitis
secca + Sjögren syndrome
1990 Saini et al. 40 PMMA CIC <1 to >1 Upper tarsal ↑ Squamous
metaplasia + GCD Decreased n.s
Biomicroscopy = papillary conjunctivitis; CIC =
Squamous metaplasia; ↑ relation in duration of
CLW
1992 Knop et al. 14 SOFT CIC ≤0.5 Upper bulbar ↑ Squamous
metaplasia + GCD Decreased 134 ± 95 Reversibility after omission of CLW
1997 Adar et al. 25 RGP, SOFT CIC ~ 3 Upper bulbar
↑ Squamous
metaplasia +
GCD; Nelson
Decreased 10.5 ± 1.1 No relation in duration of CLW
1998 Aragona et al. 86 PMMA, RGP, SOFT CIC n.s Upper bulbar Aragona Decreased n.s Less changes in RGP compared to soft lenses
2001 Albietz et al. 39 RGP, SOFT CIC ≥1 Upper, lower,
temporal Tseng Decreased n.s No relation in duration of CLW
2001 Anshu et al. 80 RGP, SOFT CIC ≤1 to >1 Upper tarsal Saini Decreased n.s More severe changes in SCL
2002 Simon et al. 28 RGP, SOFT CIC ≤0.5 Upper bulbar ↑ Squamous
metaplasia + GCD Decreased n.s Increased changes according to duration of CLW
2002 Sengor et al. 19 n.s CIC n.s Nasal and temporal
bulbar
Nelson + mapping
technique Decreased n.s
Most metaplasic changes were observed in the
lower quadrants of the bulbar conjunctiva
2006 Moon et al. 12 RGP CIC ~6 Temporal bulbar n.s Decreased n.s GCD changes attributed to CLW not keratoconic
shape of the cornea
2008 Dougthy et al 20 SOFT CIC 4-6 Nasal bulbar Nelson Decreased n.s ↓ cell area in CLWs
2008 Tomatir et al. 75 RGP, SOFT CIC 0.33-1 Upper, lower
bulbar Nelson Decreased n.s GCD related to duration of CLW
1994 Connor et al 18 SOFT CIC 0.5 Lower bulbar n.s Increased n.s GCD increased from 4 % to 7.84 %
2003 Lievens et al. 20 SOFT CIC 0.5 Lower bulbar Connor Increased 2-3% Silicone hydrogels may be slightly less irritating
2009 Corrales et al. 16 SOFT low water content;
SOFT high water content
CIC
PCR 1 Upper bulbar n.s Increased Log0.8 Water content not related to the changes
2001 Piasella et al. 14 RGP, SOFT CIC +
FC 1-20
Superior and
superotemporal
bulbar
Centrifugation +
FC
No significant
difference n.s
Fluorescence analysis showed less expression of
MUC5AC of the CL groups compared to non-
CL wearers
2006 Hori et al. 20 RGP, SOFT CIC +
FC 5-20 Temporal bulbar RNA isolation
No significant
difference
1.2 ± 0.7
(SEM)
No detectable change in mucin content between
CLW and control group
2009 Efron et al 11 SOFT LSCM 10±4 Temporal, nasal,
upper and lower LSCM
No significant
difference 111 ± 58
GCD values from CLWs higher than the control
group. No statistically significant
Abbreviations: PMMA - polymethylmethacrylate; RGP - rigid gas permeable; n.s - no specified; CIC - conjunctival impression cytology; PCR – polymerase chain reaction; FC - flow cytometry; GCD – goblet cell
density; RNA - ribonucleic acid; Log – logarithm; SEM - standard error of the mean; CLW – contact lens wear.
Literature Review 43
2.7 SUMMARY OF KNOWLEDGE GAPS AND OBJECTIVES OF
RESEARCH PROGRAM
The time-course of changes to GCD as a result of CLW is still unclear, and there is to
my knowledge, no evidence of longitudinal studies of GCD assessment using LSCM
in CLW. The work conducted to date, however, suggests CLW does alter normal
GCD in some way. It is therefore important to understand whether CLW causes a
reduction in GCD in non-CL wearers who are introduced to CLW, observe the time-
course and to determine the link between GCD and DE symptoms related to CLW.
Therefore, the primary research question of this study sought to determine the
influence of CLW on GCD in particular on CL wearers who developed DE symptoms
compared to asymptomatic wearers. This work could help researchers and
practitioners to understand the importance of these cells and the role they play in the
comfort of CLW. Longitudinal data relating GCD and CL-induced DE symptoms are
lacking, along with GCD assessed over time in CLW using LSCM.
To understand the impact and implications of external factors on ocular tissue, the
normative state must be understood. A reliable evaluation of the degree of association
between the current gold standard technique of CIC and the new, non-invasive
technique of LSCM on a healthy population has never been reported. Threfore, a
second key question of this study was the level of agreement of LSCM and CIC for
the assessment of conjunctival GCs, which attempted to fill this research gap in the
literature. Furthermore, the longitudinal interaction of these two methods was
established for the symptomatic and asymptomatic CL-induced DE and control
groups over a 6-month period. Demonstration of a direct relationship between these
two techniques will serve to validate LSCM as a viable alternative for assessing GCD
in healthy populations as well as pre-, per- and post-intervention.
To answer these two primary research questions, some aspects of the methodologies
used with LSCM and CIC needed to be addressed. Hence, prior to the longitudinal
investigation to address the two primary research questions, six studies were
conducted related to the methodological procedures.
Firstly, the presence of GCs patterns observe by LSCM have, to date, not been
confirm. It is important to understand if the assumptions of the features observed
Literature Review 44
using LSCM are indeed GCs in order to be reported and analysed. Therefore, the
research question of Chapter 3 attempted to confirm the entities presumed to be GCs
using a biopsy from conjunctival pterygium that was imaged with LSCM and stained
using antibodies for MUC5AC. It is hypothesised that GCs identify by LSCM will
indeed stained positively for MUC5AC antibody.
Secondly, the preferred staining procedure for cytological identification of GCs is
PAS stain. However, to determine GCD using CIC a more time- and cost- effective
staining procedure using Giemsa stain can be adopted. Thus, the research question of
Chapter 4 validated the cytomorphological identification for GCs to determine GCD
in samples collected by CIC stained using Giemsa stain and compared with the gold
standard cytochemical stain for mucus cells PAS. It is hypothesised that GCD
estimates from samples stained with Giemsa stain will correlate with those stained
using PAS.
Another methodological aspect to address is the reliability of LSCM to determine
GCD over time. To date, data linking intraobserver test-retest repeatability to the
consistency of GCD measurement using LSCM are lacking. Hence, the research
question of Chapter 5 demonstrated the repeatability of measuring GCD using LSCM
for a single observer on two separate occasions. It is hypothesised that LSCM is a
repeatable technique for the measurements of GCD.
Moreover, effects on GCD measurements using CIC after LSCM can occur due to a
mild epithelial cell disruption subsequent to LSCM examination. However, evidence
of this assumption is not available in the literature. Therefore, to ensure that CIC
measurements will not be influenced by the prior LSCM procedure, the research
question of Chapter 6 investigates whether prior LSCM examination compromises
GCD measured with CIC. It is hypothesised that LSCM will not influence estimates
of GCD by the prior LSCM procedure.
A random number of CIC and LSCM images have been used to determine
conjunctival GCD. Nevertheless, a sampling analysis to evaluate the minimum
number of images that accurately represent a sample for GCD has, to date, not been
described. Thus, the research question of Chapter 7 determines the minimum number
of images to estimate GCD per examination using LSCM and CIC.
Literature Review 45
Once the aspects of the methodological procedures used with LSCM and CIC were
addressed in the preliminary studies presented above, the general methodology and
research plan was developed and delimited in Chapter 8. This PhD thesis attempted to
resolve some discrepancies in the literature concerning the time course of changes of
GCD in CLW using validated methodological procedures. The development of the
methodological procedures used in these studies with regards to LSCM and CIC
allowed the examination of the two primary research questions. One of the main
research questions is answered in Chapter 9 were the level of agreement and
association between LSCM and CIC techniques for the assessment of conjunctival
GCs were examined. Finally, the primary aim of this study was address in Chapter 10
were the time course of changes in GCD in symptomatic and asymptomatic CL
wearers was investigated. The summary of these findings and recommendations for
future investigations were presented in Chapter 11.
There are some limitations related to CIC technique. For this reason, it is important to
compare this technique with a less invasive method, LSCM. CIC is a variable
technique and no evidence of repeatability of the technique has been established;
possibly because epithelial cell regrowth is needed before GC can be re-assessed.
Another limitation of this technique is that GC distribution across the conjunctiva
differs according to conjunctival region. This concept has been ignored in some
studies, where the combination of different regions of the conjunctiva (upper, lower,
temporal and nasal) have been averaged and reported. Acetate materials used for
collecting samples are also known to affect GC estimates due to variability of cell
attachment. The limitations of this technique in conjunction with other minor factors
such as lack of control for age, gender and ethnicity may play a role in the CIC
reliability. Therefore, through this study, a reasonably standardised method of CIC is
proposed for comparison with LSCM for GC assessment.
Presumed Goblet Cells Assessed by LSCM Confirmed with Immunohistochemistry in a Human Pterygium Biopsy 47
Presumed Goblet Cells Assessed Chapter 3:
by LSCM Confirmed with
Immunohistochemistry in a Human
Pterygium Biopsy
3.1 PREFACE
The structures observed with LSCM which are presumed to be GCs have, to date not
been confirmed. To study GCs using an in vivo technique, this presumption must be
explored. A novel way to confirm the nature of these cells with LSCM is to assess a
fresh biopsy firstly using LSCM and subsequently staining the sample to confirm the
nature of the cell type, in this instance, using antibodies for MUC5AC. It is known
that fixatives can induce some artefactual changes to the tissue. For that reason, to
carry out this experiment, the tissue must be a fresh biopsy and keep alive if possible
in order to have the closest possible image compared to the in vivo assessment.
However, autolytic changes can occur and be destructive to cell morphology after
several hours of 37 °C exposure. Therefore the strategy for this analysis was
performed in the shortest possible time after tissue excision.
In the human body, cell types can be distinguished from surrounding cells by their
morphological appearance at a particular tissue location. Each of these cell types has
unique antibodies that facilitate the characterization of cell phenotype which can be
analysed using methods such as immunofluorescence staining. For example,
conjunctival GCs are found scattered among the epithelial lining of the conjunctival
epithelium, having a height of three to four times that of their width and a distinct
balloon-like appearance. To date, GCs have been qualified and quantified using
LSCM based on their morphological appearance. However, for LSCM to replace CIC,
confirmation of the cell type is needed.
Characterisation of GCs in conjunctival tissue is well established, including
pterygium which is a breakdown in the normal peripheral or limbal structure and
emigration of conjunctival tissue onto the cornea. conjunctival morphology in
Presumed Goblet Cells Assessed by LSCM Confirmed with Immunohistochemistry in a Human Pterygium Biopsy 48
pterygium has been assessed previously using CIC (Chan et al., 2002) and there is no
evidence of morphological changes to the GCs in this condition. Conjunctival GCs
express a positive MUC5AC response to antibodies, squamous cells stain negatively
for cytokeratin 7 (CK-7), and GC cytoplasm is identified using CK-7 (Shatos et al.,
2003).
The in vivo evaluation of pterygium has been documented using LSCM (Zhivov et
al., 2009). The appearance of pterygium under LSCM has been described as follows.
The superficial cells are hyper-reflective with hypo-reflective borders, and the cell
nucleus is sometimes visible. The deeper cell layers are characterized by a regular
pattern of smaller cells. Capillaries with blood flow can be visualized in the stroma,
and sometimes it is possible to identify microcysts. Close to the corneal tissue,
Langerhans cells can be observed at the level of the subepithelial nerve plexus. In
some cases, a pigmented yellow line (Stocker’s line) can be visualized. Conjunctival
GCs have diameters of about 30 µm (compared to the 10 µm diameters of non-goblet
cells) with typical hyper-reflective cell bodies (Zhivov et al., 2009).
It has been demonstrated that in vivo LSCM can be used to assess GCs reasonably
well without the need for cell removal, fixation, and staining. By contrast, other
conjunctival cell assessment methods, such as CIC and biopsy, require these
procedures. LSCM shows GCs in healthy individuals to be 25 - 30 µm in diameter
(Kobayashi et al., 2005; Zhivov et al., 2006), hyper-reflective (Messmer, 2008),
bigger than surrounding cells (Villani et al., 2011a), and round (Messmer, 2008) to
oval-shaped (Pisella et al., 2001), offen with visible nuclei (Hong et al., 2010).
However, a disadvantage of this technique is the lack of identification of the specific
mucin type (i.e. MUC5AC).
GCs can also be characterised in vitro using immunohistochemestry (IHC) analysis;
this procedure enables antigen detection of cells within a tissue section. It uses
antibodies to observe a marker of interest. The principle behind IHC is the visual
demonstration of antigen-antibody binding using either a coloured histochemical
reaction or fluorescently (Taylor, 2015). However, this ex vivo technique necessitates
a long process in the laboratory, and the reagents are expensive and sensitive to light
and temperature exposure.
Presumed Goblet Cells Assessed by LSCM Confirmed with Immunohistochemistry in a Human Pterygium Biopsy 49
Based on the morphological appearance there is a fair degree of certainty that GCs
can be correctly identified using LSCM, however, it is important to use established
methods such as immunohistochemical observations of GCs in a conjunctival biopsy
to confirm this assumption at the cellular level.
3.2 PURPOSE
This study aimed to verify that the entities believed to be GCs as imaged with LSCM
are indeed GCs as confirmed by IHC, by undertaking both of these characterisation
techniques on a biopsy of excised human pterygium.
3.3 METHODS
This observational study was conducted following approval from the QUT Research
Ethics Committee and the Queensland Eye Institute (QEI) Human Research Ethics
Committee. A sample of pterygium was prepared for observation approximately 30
minutes after surgical removal from a 33-year-old male with a large nasal pterygium
and no history of ocular surgery. With the consent of the patient, a section of
approximately 800 µm² was obtained from the lower edge of the triangular- or wing-
shaped portion of the pterygium of the patient at QEI. On arrival at IHBI, the tissue
was divided into two portions and immersed in Dulbecco's Modified Eagle's Medium
(DMEM) with 5mM L-glutamine, 100 µg/ml streptomycin and 10% heat-inactivated
foetal calf serum (HI-FCS). Both samples were placed in a well plate sample holder
and transferred to the laboratory and held at 4 °C for 30 minutes during the blocking
step, which was required to prevent non-specific binding of the antibodies during the
IHC procedure.
Laser Scanning Confocal Microscopy
Using (HRT III) with the applanating surface of the TomoCap in the horizontal
position, the biopsy was placed on the centre of the TomoCap. Excess liquid from the
medium was not removed as this afforded better observation and image resolution
(Figure 3.1). The sample was carefully handled using fine surgical tweezers to avoid
tissue damage. GCs were identified and images of 400µm² were captured in steps of 1
to 2 µm deep using the section mode of the confocal microscope without moving the
tissue.
Presumed Goblet Cells Assessed by LSCM Confirmed with Immunohistochemistry in a Human Pterygium Biopsy 50
Figure 3.1 The applanating surface of the TomCap in the horizontal position, the
biopsy placed on the centre of the TomCap. Excess liquid from the medium was not
removed.
Immunohistochemistry
After LSCM observation, immunofluorescence was performed on the same section of
tissue using the double-staining method. One half of the biopsy was incubated for 2
hours at 37°C in primary antibodies anti-cytokeratin 7 (mouse anti-human CK-7;
concentration: 1mg/ml, Abcam, AU), which detects mucus-secreting cell membranes,
diluted 1:500 in phosphate-buffered saline (PBS); and antibody anti-mucin 5AC
(rabbit anti-human MUC5AC; concentration: 1mg/ml, Abcam, AU) was diluted 1:500
in PBS. After washing three times (for 2 minutes each time) in PBS using an orbital
shaker, the secondary antibodies, donkey anti-mouse fluorescein isothiocyanate
(FITC) (concentration: 2mg/ml, Abcam, AU) and donkey anti-rabbit immunoglobulin
G (IgG) tetramethylrhodamine (TRITC) (concentration: 2mg/ml, Abcam, AU) were
diluted 1:300 in PBS and the tissue was incubated for 2 hours at 37 °C. After washing
three times (for 2 minutes each time) in PBS using an orbital shaker, 1µg/ml of 4’, 6-
diamidino-2-phenylindole, dihydrochloride (DAPI) was used to determine cell nuclei,
diluted 100 times in PBS for 10 minutes. The second half of the biopsy was immersed
in 10% HI-FCS and used as a negative control (primary antibody step omitted).
High-Speed Laboratory Confocal Imaging
Prior to imaging, the tissue was placed onto a glass coverslip containing one drop of
PBS. Multiple label immunofluorescence confocal z-stack images were collected and
analysed using a Nikon A1R confocal microscope with 10x and 20x water immersion
Presumed Goblet Cells Assessed by LSCM Confirmed with Immunohistochemistry in a Human Pterygium Biopsy 51
objective lenses (Nikon Instruments Inc., Melville, NY). DAPI, FITC and TRITC
were excited with 405-nm, 488-nm, and 561-nm lasers, respectively.
3.4 RESULTS
Laser Scanning Confocal Microscopy
Using LSCM, GCs were observed in conjunctival tissue between 7 and 41 µm deep at
the level of the superficial basal cells of the pterygium tissue. In the sample, some
GCs were brighter than others, possibly due to overlapping cells. This suggested that
GCs were located at different depths in the tissue. GCs were estimated to be
approximately 20 to 30 µm in diameter, although GC diameter varied according to
depth in the tissue. GCs appeared to have a smaller diameter in the deeper layers (35
to 40 µm from the surface). Round GCs were smaller in diameter than oval-shaped
GCs. A small dark dot was visible in some GCs, potentially indicating the nuclei, or
perhaps the opened apical portion, indicating the site of mucin release. GCs were also
more reflective and bigger than the surrounding cells. A diagonal hyper-reflective line
was seen across the sample in Figure 3.2 A; this may have been an artefact of the
processing or a fold of the tissue sample. GCs were more distinct and more dense in
the lower part of the field, possibly due to non-uniform thickness and therefore slight
defocus of the regions of the surface.
Immunohistochemistry
Positively stained GCs in immunofluorescence showed a similar distribution pattern
to those observed during the LSCM examination. In the tissue sample, GCs exhibited
intense staining for CK-7 (Figure 3.2 A). The tissue sample also stained intensely for
the GC-specific mucin type, MUC5AC (Figure 3.2 B), whereas deoxyribonucleic acid
(DNA) expression was indicated by strong binding to DAPI in the nuclei of all cells
(Figure 3.2 C). Immunofluorescent staining with the isotype control using secondary
antibodies showed no apparent immunoreactivity (Figure 3.2 D). Thus the recognition
of non-goblet cells was also possible by overlapping the images as shown in Figure
3.3 B.
Presumed Goblet Cells Assessed by LSCM Confirmed with Immunohistochemistry in a Human Pterygium Biopsy 52
FITC TRITC DAPI CONTROL + DAPI
Figure 3.2. Immunolocalization of goblet cell markers in a pterygium biopsy (A)
goblet cell cytoplasm showed a band of CK-7 using FITC conjugated (green). (B)
location of mucin expression MUC5AC was labelled with TRITC (red). (C) 4, 6,
diamidino-2-phenylindole (DAPI) to identify nuclei (blue). (D) anti-mouse and anti-
rabbit isotypes control. Magnification 100X.
In order to analyse cell morphology, a 20x objective lens was used and a section of
200 µm² was taken from Figure 3.3 B and represented in Figure 3.3 C. Three different
GC appearances were recognized as shown in Figure 3.3 C. The bigger cells (oval-
shaped) seemed to have more MUC5AC expression and absent nuclei, whereas the
smaller cells (round) showed distinct apical nuclei and less MUC5AC expression.
Another appearance of GCs identified in the sample indicated the absence of
MUC5AC and the distinct balloon-like appearance. However, the edges of the nuclei
outlines were visible, indicating that they were positive for CK-7.
Secondary antibody binding was examined in the negative control and resulted in a
negative signal, indicating that there was not contamination and non-endogenous
labelling (Figure 3.2 D).
Presumed Goblet Cells Assessed by LSCM Confirmed with Immunohistochemistry in a Human Pterygium Biopsy 53
Figure 3.3. Characterization of goblet cells from pterygium biopsy using laser scanning confocal microscopy and immunohistochemistry. (A) in
vivo LSCM image shows distinct balloon-like cell appearance (yellow arrow) compared to the squamous non goblet cells (dotted red arrow) (B)
Immunofluorescence image of same tissue from image (A) triple-labelled using FITC+TRITC+DAPI. The dotted arrow represents positive stain
for GC and solid arrow represents negative stain and presence of nucleus assumed to be squamous non-goblet cells. (C) two times magnification
from B showing three distinct cell types with positive CK-7. Dashed arrow represents large and oval-shaped GCs with positive MUC5AC
expression (red). The dotted arrow represents smaller and round-shaped GCs with less intensive red than the larger cells. The solid line represents
possibly immature GC lacking the balloon-like appearance and MUC5AC expression.
Presumed Goblet Cells Assessed by LSCM Confirmed with Immunohistochemistry in a Human Pterygium Biopsy 54
3.5 DISCUSSION
The study was designed to image and characterise a single piece of human
conjunctival tissue containing GCs using both LSCM and immunofluorescence. This
study has successfully imaged and identified, for the first time, GCs in conjunctiva
using LSCM and IHC using a human pterygium biopsy. The intention to observe an
identical portion of tissue was not precisely realised due to manipulation of the tissue
and modification of the tissue by the staining procedure. It is reasonable, to assume
however that the structures observed in the tissues by LSCM were in fact GCs despite
the slight mismatch in the location observed of the tissue.
Morphological analysis with LSCM allowed the identification of goblet and epithelial
cells. In some parts of the tissue, epithelial cells were observed to have a distinct cell
membrane with bright or dark spots. The cytoplasm was observed in some cells with
low contrast. When visible, the nucleus was dark and round, but was difficult to
identify in most of the cells. This was possibly due to the poor resolution of the small
nuclei (approximately 3 to 5 µm). GCs appeared bright, were sometimes round to
oval-shaped, and demonstrated rich contrast. They were considerably larger than non-
goblet cells with defined borders and varied sizes, between 20 and 30 µm in diameter.
These observations were also confirmed in healthy individuals by other authors using
the same technique (Rath et al., 2006; Villani et al., 2011a; Zhu et al., 2009).
LSCM has the advantage of allowing in vivo assessment of GCs at different depths in
the epithelium in vivo, and has the capability of assessing cells before and after
interventions without the need for tissue removal and sample processing. A limitation
of this technique is the relatively small field of view (400 x 400 µm) and the fixed
observations through the 60X objective lens. In comparison, the fluorescein and
confocal microscopes provide varied magnifications (10X, 20X, 40X, and 60X).
The IHC method has been used to determine different mucin types, which are not
only expressed by the GCs but also by the squamous cells of the conjunctiva (Rios et
al., 1999). However, in the biopsy observations, three distinct cell morphologies were
identified as GCs based on positive staining for MUC5AC and/or CK-7, as well as the
morphologically distinct balloon-like appearance. Many authors have extensively
described and analysed the round and oval-shaped cells. Likewise, in this study, the
Presumed Goblet Cells Assessed by LSCM Confirmed with Immunohistochemistry in a Human Pterygium Biopsy 55
presence of MUC5AC expression was shown to be more intense in the oval-shaped
cells than in the round ones. This was probably attributable to a higher number of
mucin granules inside the GCs. In the magnified portion of the tissue shown in Figure
3.2C, it was interesting to observe that a few nuclei were positively outlined with CK-
7 but lacked MUC5AC expression and balloon-like morphology. A possible
explanation for this observation was that some single GCs from the conjunctiva have
not fully differentiated. Cells of this nature have been observed in rabbit and human
cell culture studies (Hodges and Dartt, 2010). Another possible reason for this
staining pattern is that these cells represented either degenerating limbal epithelial
cells in the pterygium, or morphologically altered cells resulting from tissue damage
due to the surgical excision process or laboratory processing.
This experiment gives confidence in future observations specifically in this
longitudinal study because the presumed GCs assessed by LSCM were also observed
on the same tissue using confirmatory processes. Assumptions of GCs visualized with
confocal are based on shape size and reflectivity of the cells. The shape, size and key
identifying factors were also observed in vitro.
In summary, presence of GC patterns were observed by LSCM and IHC, in a single
human biopsy of pterygium providing adequate evidence to conduct the longitudinal
observations of GCs using both in vivo and ex vivo cytological techniques.
Validation of Giemsa Stain using PAS for Goblet Cell Density Assessment 57
Validation of Giemsa Stain using Chapter 4:
PAS for Goblet Cell Density
Assessment
4.1 PREFACE
The most common stain for mucus detection using an immunocytochemical non-
fluorescent approach is PAS. This stain has been used and recommended for the
detection of carbohydrate macromolecules such as glycogen, glycoprotein, and
proteoglycans, usually located in connective tissues like mucus, the glycocalyx, and
basal laminae. Conjunctival samples obtained by CIC are usually stained with PAS
for detection of GCs and are considered as positive mucous cells for PAS. This
staining does not highlight the nucleus or cell walls. However, the balloon-like shape
can be seen in a pink colour and counterstained purple background by using combined
stains such as hematoxylin.
PAS has some limitations when assing GCs on cytological samples. For example,
some studies have suggested that the borders of GCs using PAS are difficult to
observe (Albietz et al., 2003; Murube and Rivas, 2003; Rolando et al., 1990). These
reports can possibly be attributed to inconsistent technique during sample collection
and fixation. Furthermore, sometimes two to three layers of epithelial cells are needed
in order to give support to the weight and size of the GCs attached in the filter used
for CIC. Sometimes GCs can also be slightly overlapping in samples with high GCD
(Doughty, 2012b). For this phenomenon, GCs can be miscounted when calculating
GCD using PAS. Also, using PAS for GC estimates could lead to false positive
findings due to mucin content being released which is not a component of a cell.
Another stain used for detection and density calculation for cytomorphological
findings in CIC is Giemsa. This stain is used to differentiate nuclear and/or
cytoplasmic morphology of certain cell types. It highlights phosphate groups of DNA,
especially in regions with high amounts of adenine-thymine. Giemsa stain is also used
to visualize chromosomes that stain magenta colour and detects parasite bodies which
turn pink. This stain does not highlight mucous but rather facilitates the identification
Validation of Giemsa Stain using PAS for Goblet Cell Density Assessment 58
of GCs which appear as distinct balloon-like cells with smaller-sized nucleus than
surrounding cells. Non-goblet cells stain pale-blue in the cell walls with a dark-blue
nucleus. As mentioned previously, overlapping GCs can be miscounted when
quantifying GCD. The presence of a distinctive nucleus and cell outline makes cell
counting more consistent and minimises the potential for miscounting cells.
4.2 PURPOSE
The purpose of this study was to compare the use of PAS and Giemsa stains to assess
GCD in CIC samples.
4.3 METHODS
This study was conducted following approval from the QUT Research Ethics
Committee. To determine GCD using Giemsa and PAS staining procedures for 10
healthy volunteers from IHBI underwent CIC on one occasion after signing informed
consent. Participants first underwent slit-lamp biomicroscopy examination of the
ocular surface to ensure ocular and conjunctival integrity. The methodological sample
collection using CIC was performed as described in Chapter 7 (Section 7.2.1 ‘Sample
collection’).
After no more than 24 hours of fixation in 95% methanol, the samples were divided
into two equal parts. One-half was stained with Giemsa stain as described in Chapter
7 (Section 7.2.2 ‘Staining procedure’). The second-half was stained with PAS using
the following staining protocol. To avoid sampling bias, the examiner was masked to
the identity of the participant and to which half of the filter was stained with Giemsa
or PAS.
Staining Protocol:
1. Rehydration
70% ethyl alcohol 2 min
Tap water 10 dips x 2
2. PAS
a) Periodic acid 0.5%
2 min
Tap water
10 dips x 2
Validation of Giemsa Stain using PAS for Goblet Cell Density Assessment 59
b) Schiff reagent 1:3
freshly diluted with
distilled water
2 min
Tap water 10 dips x 2
c) Sodium metasulfite
0.5%
2 mm
Tap water 10 dips x 2
3. Gill's hematoxylin 1 min (or 4 min if freshly
made)
Tap water 10 dips x 2
4. Scott's tap water
substitute
2 min
Tap water 10 dips x 2
5. Dehydration 95 % ethyl alcohol 10 dips x 2
6. Modified OG-6
2 min
95% ethyl alcohol 3 min
7. Modified EA 2 min
95 % ethyl alcohol 10 min
8. Dehydration Absolute alcohol 5 min
9. Transfer to xylene 15 min
After the staining procedure, each half was mounted in separate glass slides for
observation under the microscope as shown in Figure 4.1
Validation of Giemsa Stain using PAS for Goblet Cell Density Assessment 60
Figure 4.1. Half of the CIC specimen stained using PAS (A) and the second half
stained using Giemsa (B).
The image capture procedure used is that described in Section 7.2.3 ‘Image capture’
of Chapter 7, and the image selection criteria applied for Giemsa staining was that
described in Section 7.2.4 ‘Image selection criteria’ of Chapter 7. For PAS, the
following image selection criteria was applied.
Image selection criteria
Images from CIC with no disrupted cell material which contained GCs approximately
25 to 30µm in diameter were selected. The GCs had a pink membrane and were
easily differentiated from surrounding cells because of their balloon-like appearance
and cell size. (Figure 4.2)
Figure 4.2. Conjunctival impression cytology sample of nasal bulbar conjunctival
stained with PAS and counterstained with Gill’s haematoxylin stain. (magnification
200X). The white arrow points to an epithelial cell and the black arrow points to a
Validation of Giemsa Stain using PAS for Goblet Cell Density Assessment 61
goblet cell.
Determination of GCD using both Giemsa and PAS samples was used as describe in
Chapter 7 (Section 7.2.6 ‘Determination of goblet cell density’).
The paired sample t-test was used to determine differences between GCD obtained by
the two samples from the same participant. Pearson’s r was used to determine the
correlation between the two variables. IBM SPSS Statistics V21 was used for the
analysis.
4.4 RESULTS
Normal distribution was determined using Shapiro-Wilcoxon test. Paired t-test
revealed no significant difference between Giemsa and PAS staining procedures for
GCD (p = 0.64). The mean difference was 7.8 cells/mm², and Pearson’s correlation
was 0.58 which indicates a strong positive correlation between the two staining
methods.
Table 4.1. Basic descriptive statistics for GCD collected by CIC technique using
Giemsa and PAS staining procedures. Values are presented in mean ± SD or count for
categorical variables.
Characteristics of the participants Range Age (mean ± SD) 29 ± 9 18 - 50
Gender (male/female) 4 / 6 - Goblet cell density (cells/mm²) Giemsa PAS
Average (mean ± SD) 410 ± 66 418 ± 58 Min - Max 291 - 519 301 - 503
N 10 10
Validation of Giemsa Stain using PAS for Goblet Cell Density Assessment 62
Figure 4.3. Bland-Altman plot of the relationship between differences in GCD
obtained by Giemsa and PAS staining procedures vs. GCD mean. The middle line
represents the mean difference between the two measurements (7 cells/mm²). The
upper and lower lines (dashed) represent the 95% LoA, +116 (upper bound) and -109
(lower bound) including 0. There are 10 data points (1 per participant) that represent
the difference between Giemsa and PAS. Each data point represents the average value
of GCD in 5 images at each testing time. One outlier is observed in the plot.
4.5 DISCUSSION
In this study GCD measurements using Giemsa and PAS staining procedures showed
no statistically significant difference (p = 0.64) and a positive correlation (r = 0.58).
This finding indicate that cytomorphological identification of GCs to estimate GCD
using Giemsa stain can be used in place of the immunocytochemical mucus detection
by PAS.
Staining for mucin detection with PAS has probably been the first option of previous
studies in order to detect GCs in cytological samples because this stain highlights the
presence of mucus as a pink colour and the background is counterstained with a
pourple colour. However, Giemsa stain has the advantage of facilitating visualization
of cell borders and cell nuclei, thus making cell counting more reliable.
In this experiment, each PAS sample required a processing time of 1 hour and 20
minutes (13 hours and 30 minutes in total) and the total cost of reagents was
Validation of Giemsa Stain using PAS for Goblet Cell Density Assessment 63
approximately AUD 1000, excluding laboratory glassware and Millicell inserts. In
contrast, Giemsa stain required a processing time of 45 minutes per sample and the
reagents cost approximate AUD 120. Thus, Giemsa staining has the additional
advantage over PAS staining as being time- and cost-effective.
In conclusion, this study validates the cytomorphological identification for GCs to
determine GCD using Giemsa stain with the histochemical mucus detection by PAS.
Also, Giemsa stain is less expensive and provides faster results than PAS. Therefore
in the main longitudinal study, identification and quantification of GCs obtained using
the CIC technique was performed using Giemsa stain.
Repeatability of Measuring Goblet Cell Density Using LSCM 65
Repeatability of Measuring Goblet Chapter 5:
Cell Density Using LSCM
5.1 PREFACE
This Chapter presents an experiment which sought to determine the intra-observer
test-retest repeatability of the GCD using LSCM. The reliability of these
measurements is used to determine the consistency of the technique over time.
Only one report was found in the literature where repeatability was analysed for GCD
of the inferior tarsal conjunctiva measured using LSCM (Villani et al., 2011a). Inter-
observer measures (two observers) were significantly different (362 ± 399 cells/mm²
and 634 ± 365 cells/mm² for observer 1 and 2, respectively) and were not correlated.
These investigators concluded that the test was not repeatable for assessing GCD
using multiple observers.
This study was designed to assess intra-observer test-retest repeatability of the GCD
of the nasal bulbar conjunctiva. Image cell counts were analysed using the cell count
mode of the HRT III software (Cell Count Software; Heidelberg Engineering GmbH).
There were no reports found in the literature regarding test-retest intra-observer
repeatability of GCD measured using LSCM.
5.2 PURPOSE
To assess the intra-observer repeatability of GCD from the nasal bulbar conjunctiva
using LSCM.
5.3 METHODS
This study was conducted following approval from the QUT Research Ethics
Committee. To assess the consistency of GCD measurements, 10 healthy participants
underwent LSCM of the nasal bulbar conjunctiva at two different times by a single
observer after signing informed consent. The test was conducted on two separate
occasions of at least 2 days apart. Prior to the examination, participants underwent
slit-lamp biomicroscopy examination of the ocular surface to ensure conjunctival
Repeatability of Measuring Goblet Cell Density Using LSCM 66
integrity.
The instrument focal plane was set at depth of 10 µm, and the centre of the front
surface (TomoCap) of the instrument was placed at approximately 2 to 4 mm from
limbus area. The applanating lens was displaced slightly in vertical and horizontal
movements while the focal plane was gradually focused on subconjunctival tissue
with the aim of capturing different GC groups. For image analysis, three high quality
frames not overlapping by more than 20% were selected for each examination. For the
count of GCs, the definition of GCs was essentially consistent with those reported
previously in the literature (Efron et al., 2010a; Hong et al., 2010; Kobayashi et al.,
2005; Kojima et al., 2010; Villani et al., 2011a; Wei et al., 2011; Zhivov et al., 2006;
Zhu et al., 2010). Descriptions of GCs morphology by Messmer (2006) and Rath
(2006) vary considerably compared to the authors mentioned above because they are
more consistently with conjunctival microcyst. Therefore these descriptions were
excluded for GC identification using LSCM. For the cell counts, the average cell
count of three images was applied per examination.
Description of GCs using LSCM: GCs were approximately 30 µm in diameter, hyper
reflective, bigger than surrounding cells, round to oval in shape and sometimes had a
visible nucleus, as shown in Figure 5.1.
Figure 5.1. In vivo confocal image of nasal bulbar conjunctival of goblet cells. The
white arrow points an epithelial cell and the blue arrow points a goblet cell.
Repeatability of Measuring Goblet Cell Density Using LSCM 67
Assessment of intra-observer repeatability was conducted using the interclass
correlation coefficient (ICC), limits of agreement (LoA) and the Bland-Altman plot
(Bland and Altman, 1986). A one-way random effects model was used to test for
consistency of the measurements for the participant group. IBM SPSS Statistics V21
was used for this analysis.
5.4 RESULTS
Demographic characteristics and GCD values of participants are shown in Table 5.1.
The paired t-test was used to compare test-retest measurements. No significant
difference between test-retest values of GCD using LSCM was found between the
examinations (p = 0.05). The mean difference between test-retest was 14.30
cells/mm². ICC value for GCD demonstrated adequate repeatability (ICC 0.76, 95%
CI = 0.11 – 0.94). The results of correlation, ICC and LoA are shown in Table 5.2. A
Bland-Altman plot is also shown in Figure 5.2
Table 5.1. Demographic characteristics of participants of the intra-observer
repeatability analysis of GCD measures using LSCM. Values are presented in mean ±
SD or count for categorical variables.
Parameter Range
Age (years) 34 ± 6 25 - 46
Sex (male/female) 6/4 -
GCD (cells/mm²)
Test 404 ± 84 295 - 555
Retest 390 ± 113 240 - 591 GCD, goblet cell density
Table 5.2. Mean difference, ICC and LoA of the intra-observer repeatability analysis
of GCD measures using LSCM.
Mean difference
(test – retest) ICC
95% CI LoA
Lower Upper Lower Upper
GCD (cells/mm²) 14 0.76 0.11 0.94 -162 +191
GCD, goblet cell density; ICC, interclass correlation coefficient; CI, confidence interval; LoA
limits of agreement
Repeatability of Measuring Goblet Cell Density Using LSCM 68
Figure 5.2. Bland-Altman plot of intra-observer test-retest of GCD. Relation between
differences in GCD vs. GCD mean. The middle line represents the mean difference
between the two measurements (14 cells/mm²). The upper and lower lines (dashed)
represent the 95% LoA, +191.87 (upper bound) and -162.41 (lower bound) including
0. There are 10 data points (1 per participant) that represent the difference between
test and retest. Each data point represents the average value of GCD in 3 images at
each testing time.
5.5 DISCUSSION
In this study, the repeatability of GCD measurements within one observer was
examined. The test conducted on two separate occasions of at least 2 days apart
showed an acceptable level. The ICC of 0.76 and the coefficient of variation of 12%
indicate a reasonably acceptable repeatability level with low percentage of variation
given the difficulty of this technique.
Popper et al. (2003) performed repeated measurements of the epithelial cells of the
cornea of 20 healthy participants on two occasions separated by 14 days using LSCM
and demonstrated inter-observer repeatability to be 5.8% in cell densities of the
cornea. Also, epithelial cells densities of the bulbar conjunctiva measured with LSCM
cell count mode demonstrated intra-observer coefficient of variability of 3.2% on the
average of the four cardinal points of the bulbar conjunctiva (Efron et al., 2009).
Although, these two studies did not report GCD repeated measures, they
Repeatability of Measuring Goblet Cell Density Using LSCM 69
demonstrated high repeatability of measurements of different cell type of the ocular
surface using LSCM.
Variations in GCD measurements may occur because LSCM only captures an area of
400 x 400 µm², therefore, finding the same area in the bulbar conjunctiva in the
second measurement can be affected. Further, participant cooperation and eye
movements do, in part, influence the variability of these measurements.
Values of GCD were similar to those of Zhu et al. (2009), and Hong et al. (2010) who
obtained an average of 423 ± 70 cells/mm² from the four cardinal points of the bulbar
conjunctiva and 332 ± 137 cells/mm² from the superior bulbar conjunctiva;
respectively, using LSCM. Although this experiment determined with 95%
confidence the measurement of GCD was within ± 175 cells in the repeated test, this
is considered to be a reasonably high level of variability and conclusions should be
treated with a degree of caution. Therefore, studies with a higher number of
participants should be conducted in the future.
There is a lack of evidence in the literature related to stage and time course of
conjunctival GCs and what factors may influence observations in vivo. However,
animal model and ex vivo human studies suggest that temperature, osmolarity and
nerve stimulation increase mucin production, and GC proliferation may be regulated
by genetic programming (Hong et al., 2010). It is also possible that GCs observed by
LSCM are shown as holes representing the cell in the process of expelling the mucin
content (Efron et al., 2009). When calculating GCD in this experiment description of
presume GC was based on previous studies. However, there could be a cycle of the
GC (mucus production, synthesis or secretion) where the in vivo appearance (dark or
bright) may influence the visualization with LSCM, leading to errors in in the density
calculation.
Inter-observer repeatability was not assessed due to lack of resources and logistical
constraints. However, since only a single observer (the candidate) was used for all
experiments described in this thesis, a determination of intra-observer repeatability
was of prime importance.
Repeatability of Measuring Goblet Cell Density Using LSCM 70
In conclusion, the repeatability of measuring GCD using LSCM was demonstrated for
a single observer. The results indicated acceptable repeatability for the purposed of
the proposed longitudinal study (see Chapter 10). Inter-observer repeatability for
determining GCD using LSCM would need to be established for future studies
conducted by two or more observers.
Effect of Test Order on Goblet Cells Assessment 71
Effect of Test Order on Goblet Chapter 6:
Cells Assessment
6.1 PREFACE
Although LSCM is considered as a non-invasive procedure, some epithelial disruption
can occur by the TomoCap. There is a mild positive staining after LSCM evaluation
on the ocular surface which indicates a minimal debridement is caused by the contact
of the TomoCap with the surface of the eye. This suggests the technique may impact
on the outcome of any subsequent tests. Performing LSCM first is the only option
since CIC by its nature alters the available cell population. Therefore, a literature
search was conducted to find evidence of GCD assessed by CIC after LSCM
examination. The use of CIC after LSCM assessment on the same area of examination
(bulbar conjunctiva) has been implemented to compare CIC and LSCM to measure
GCD related to epithelial cell changes (squamous metaplasia) in ocular surface
disorders such as those caused by glaucoma treatment (Ciancaglini et al., 2008),
Sjogren’s syndrome (Hong et al., 2010), DE (Kojima et al., 2010), pterygium (Labbé
et al., 2010), chemical burns (Le et al., 2010), tafluprost therapy for glaucoma
(Mastropasqua et al., 2013) and atopic keratoconjunctivitis (Wakamatsu et al., 2009).
None of these studies revealed if GCD assessed by CIC was affected by previous
assessments using LSCM. The duration of time between the use of these techniques is
unclear with the exception of Mastropascua et al. (2013), who separated the LSCM
and CIC measurements by 24 hours.
The evaluation of GCD on the central cornea and the nasal and temporal bulbar
conjunctiva of patients with ocular chemical burns has revealed similar cell counts for
LSCM and CIC (136 ± 79 and 121 ± 66 cells/mm², respectively) and show a positive
correlation (Spearman’s rho ρ = 0.92; p < 0.001) (Le et al., 2010). This may indicate
that the procedure of LSCM examination does not casue GC loss. The following study
was designed to investigate whether GCD measured with CIC is compromised by
conductng LSCM before CIC.
Effect of Test Order on Goblet Cells Assessment 72
6.2 PURPOSE
To investigate whether GCD measured with CIC is compromised by prior LSCM
examination.
6.3 METHODS
This study was conducted following approval from the QUT Research Ethics
Committee. To investigate the impact on a specific testing order on GCD of the nasal
bulbar conjunctiva, LSCM and CIC were performed on nasal bulbar conjunctiva of
the right eye and to serve as a control, CIC only was performed in the same location
of the left eye in 10 healthy volunteers after signing informed consent. Prior to the
examination, participants underwent slit-lamp biomicroscopy examination of the
ocular surface to ensure conjunctival integrity. The methodological GC assessment
using LSCM and CIC were performed as described in Chapter 7. For the purpose of
statistical analysis, ten CIC measures of GCD from right eyes (GCD LSCM+CIC)
were compared to ten measures from the fellow eye for the same participant (GCD
CIC).
This methodology was adopted because the literature suggests that there is a high
correlation between GCD of the right and left eyes at all four cardinal points of the
bulbar conjunctiva using CIC (Morales-Fernández et al., 2010).
The following technique was used for CIC. A few minutes after performing LSCM as
described in Chapter 7, the right eye was anaesthetized again and the centre of a
Biopore membrane (Millicell cell culture inserts; Millipore Corp, Cork, Ireland,
United Kingdom) was gently applied to the nasal bulbar conjunctival surface at
approximately 2 to 4 mm from the limbus area. The sample was allowed to air dry
and then immersed in 95% methanol for fixation using a well culture plate sample
holder. A second sample from the left eye was taken from the same location (nasal
bulbar conjunctiva). The samples were refrigerated at 4 ºC for no more than 24 hours.
To verify the location of the impression and the integrity of the exposed bulbar
conjunctiva, a slit lamp examination with fluorescein was conducted under cobalt blue
illumination with a yellow Boston filter.
The staining procedure of the sample was performed using Giemsa stain according to
Effect of Test Order on Goblet Cells Assessment 73
the following guidelines from the manufacturer (Sigma-Aldrich): Millicell inserts
with approximately 60% of cellular material across the field of the filter were
assessed. The same well culture plate sample holder was used to retain the specimens
during staining. The specimen was allowed to air dry at room temperature; the
Giemsa stain was diluted 1:20 with deionized water and the specimen was immersed
in the diluted Giemsa solution for 30 minutes. The sample was rinsed with tap water
prior to examination.
Images of the conjunctival sample were captured by a Leica DM2500 microscope
(Leica Microsystems) to visualize the cells collected. This system had a magnification
of x200 and field of view of 640 x 480 µm². Cytomorphological identification of GC
using Giemsa stain was undertaken according to the image selection criteria described
below. Five images were used to determine GCD from each sample. A validation of
the image analysis approach for CIC was carried out (see Chapter 7) and the number
of images used to determine GCD from each sample changed in the main longitudinal
study.
Description of GCs using CIC: Images from CIC with confluent cell material
contained GC approximately 25 - 30 µm in diameter as shown in Figure 6.1. The cells
had a pale membrane with defined borders, and a visible nucleus localised centrally,
although sometimes eccentrically in bigger cells (approximately 30 µm) (Doughty,
2011a). GCs were easily differentiated from surrounding cells because of their
balloon-like appearance and cell size (Doughty, 2012a).
Effect of Test Order on Goblet Cells Assessment 74
Figure 6.1. Conjunctival impression cytology of nasal bulbar conjunctival stained
with Giemsa stain (magnification 200X). The white arrow points to an epithelial cell
and the black arrow points to a goblet cell.
A paired sample t-test was conducted to compare significant effects on GCD
measures using CIC method of assessment prior LSCM.
6.4 RESULTS
The average GCD (LSCM before CIC) and GCD (CIC) were 461 ± 126 cells/mm²
and 431 ± 168 cells/mm², respectively. There was no significant difference between
the two main outcome variables (p = 0.28). There was also a high correlation between
GCD (LSCM before CIC) and GCD (CIC), Pearson's correlation (r = 0.87), as shown
in Table 6.1
Table 6.1. Descriptive statistics for GCD assessed by CIC with and without prior
LSCM and paired sample t-test between the two main outcome variables (OD =
LSCM prior GCD (CIC)) and (OS = GCD (CIC)) Values are presented as mean ± SD.
LSCM before CIC CIC
N 10 10
GCD (CIC) (cells/mm²) 461 ± 126 431 ± 168
N, number of eyes; GCD, goblet cell density; LSCM; laser scanning confocal microscopy; CIC, conjunctival impression cytology; CI, confidence interval.
Mean difference 95% CI
r p value Upper Lower
GCD (cells/mm²) 30 ± 83 -29.49 89.89 0.87 0.28
Effect of Test Order on Goblet Cells Assessment 75
6.5 DISCUSSION
The purpose of this pilot study was to demonstrate whether GCD assessed using CIC
was affected by prior LSCM examination. This small study demonstrates that the
assessment of GCD using LSCM before CIC does not affect GC counts assessed by
CIC despite the mild debridement of superficial epithelial cells evident with
examination with sodium fluorescein after using LSCM. This could be attributed to
the need for active GCs to be attached to the surrounding tissue (at least two to three
cell layers) for support to be removed from the conjunctival surface (Doughty, 2012a)
which maybe LSCM only causes a mild debridement of a few cells on the surface
loosely attached, and ready to detach or slough off the surface.
Mastropasqua et al., 2013 considered the possible effect of LSCM on GCD when
conducted prior to CIC in participants treated with preservative free therapy for
glaucoma and recommended a period of at least 24 hours be allowed to avoid
misinterpretation of CIC results for the GC counts (Mastropasqua et al., 2013). This
recommendation, however, was anecdotal and not supported statistically.
The GCD in healthy participants ranges from 380 to 620 cells/mm² according to
researchers that have used the CIC technique (Adams et al., 1988; Doughty, 2012a;
Nelson, 1988; Tseng et al., 2001; Tseng, 1985). However, due to the invasive nature
of this technique where cells are removed, repeated sampling results in a reduced
GCD using the same sampling area (Rolando et al., 1994). In contrast, the
repeatability of measuring GCD using LSCM was previously demonstrated in the
previous intra-observer study in Chapter 5, which showed acceptable repeatability
after 2 days of assessment in the same area of the bulbar conjunctiva. These
assumptions suggest that CIC is not only less repeatable in a short time period but is
also a more invasive technique than LSCM.
In conclusion, this study demonstrated that there are no significant effects on GCD
measured using CIC after LSCM examination. Therefore, in the main longitudinal
study a GC assessment was performed using CIC a few minutes after the LSCM
examination.
Image Sample Size for Goblet Cell Density Determination 77
Image Sample Size for Goblet Cell Chapter 7:
Density Determination
7.1 PREFACE
The determination of GCD using image processing assessments is subject to a variety
of challenges during sample acquisition (e.g the cooperation of participants for sample
collection), sample processing, and reproduction of the same techniques when
collecting longitudinal data. Any discrepancy in this processing may result in a
degradation of poorly defined methods in research studies. For applications in which
images are ultimately to be reliable in a longitudinal study, an image sampling and
processing assessment needs to be conducted to ensure an acceptable level of
precision for the GC estimates using LSCM and CIC.
Sampling images from techniques that are not fully standardized for the assessment of
GCD is a worthy process to be defined in the methods of this investigation. However,
it could be argued that GCD could be biased by the influence of pathological
conditions. For example, there may be factors that should be taken into account that
might increase the variability in the GCD of a pathological eye, such as characteristics
of the image that is selected for the cell count. In other words, the images from a
healthy individual are usually images that are covered in cell material with an
abundant number of GCs compared to images from an individual with an ocular
surface condition (i.e. Sjögren's Syndrome dry eye). However, this could also be
attributed to difficulties during sample collection such as cell attachment to the filter
using CIC or poor focusing and eye contact between the TomoCap and the
conjunctival surface during LSCM. Also, a pathological condition may alter the
morphology of the cell. This could influence the cell count because the GC could be
hard to distinguish and be miscounted, especially using LSCM because of the slightly
lower resolution and the grey scale imaging.
The importance of determining an image sample to calculate GCD in healthy
individuals is explained in this Chapter. However, it is important to acknowledge that
Image Sample Size for Goblet Cell Density Determination 78
this image sampling technique may need to vary in pathological conditions and
further investigation should address these factors in order to minimize variability in
the calculation of GCD when investigating ocular pathologies.
A random number of CIC and LSCM images have been used to determine
conjunctival GCD. This suggests that previous reports using any of these two
techniques for CGs assessment lack analytical support to determine the minimum
number of images selected for the GCD estimates.
Some investigators that have used CIC to estimate values of GC counts have selected
up to 10 images from a single sample to report the average of GCD on multiple
locations of the conjunctiva (Murube, et al., 2003; Paschides, et al., 1991; Rivas, et
al., 1995). However, there is no evidence in any of these reports about the strategy
used to determine a representative number of images selected for the GCD estimates.
Other limitations of these studies are (a) the lack of information in regards the quality
of the image to consider for GC counts, and (b) whether the images were overlapping.
To address these issues, a standard protocol with clear criteria was established to
determine GCD.
Various sampling strategies have been applied to LSCM to quantify aspects of the
ocular surface. Vagenas et al., (2012) used a random sampling paradigm to evaluate
the number of images to use as a representative sample for analysis. However, there is
no evidence of such analysis to accurately determine GCD using LSCM.
To ensure an acceptable level of precision for GCD calculation using CIC and LSCM,
an image sampling investigation was conducted.
7.2 PURPOSE
The purpose of this analysis was to determine the minimum number of images for
GCD calculation using CIC and LSCM.
Image Sample Size for Goblet Cell Density Determination 79
7.3 METHODS
In order to determine the minimal number of images for GCD calculation using CIC
and LSCM the following identification and image criteria strategies were applied and
a validation for an image sample approach was analysed.
7.3.1 Conjunctival Laser Scanning Confocal Microscopy
Goblet cell identification
No evidence could be found in the literature about the morphological differences
between GCs from CL wearers and healthy non-CL wearers individuals observed
under LSCM. However, previous observations of GCs using LSCM on healthy
participants have used different terms to describe these morphological features
(shown in Table 7.1). GCs have been described to be homogeneous in brightness
(Kobayashi et al., 2005; Wei et al., 2011), hypo-reflective cells (Messmer et al., 2006)
paler than surrounding cells (Efron et al., 2009) and hyper-reflective cells
(Mastropasqua et al., 2013; Messmer, 2008).
In terms of diameter, the authors have agree to the dimensions described by Kessing
in 1968 who reported variations in cell size between 25 to 30 µm in diameter
(Kessing, 1968). This variation in diameter could be attributed to mucin content and
cell cycle. The shape of the GCs under LSCM is reported to be round (Messmer et al.,
2006); roundish (Villani, Beretta, Galimberti, Viola & Ratiglia, 2011) and oval in
shape (Mastropasqua et al., 2013). The variety in the descriptions of the GCs
observations under LSCM lead to unify some of the previous descriptions into the
most appropriate for this study based on the images already obtained from healthy
participants in the repeatability study in Chapter 5. The features that will be counted
as GCs are hyper-reflective cells (Messmer, 2008), bigger than surrounding cells;
approximately 25-30 µm in diameter (Kobayashi et al., 2005; Zhivov, Stachs, Kraak,
Stave & Guthoff, 2006) and round to oval in shape (Mastropasqua et al., 2013;
Messmer, 2008), sometimes with visible nucleus. This description excludes reports by
Messmer and Rath 2006 because these features have the appearance of conjunctival
microcysts.
Image Sample Size for Goblet Cell Density Determination 80
Table 7.1. Previous observations of goblet cells under LSCM.
Author / Goblet cells image
obtained by LSCM
Morphological goblet cell description by LSCM in
healthy participants
(Kobayashi et al., 2005)
Presumed GCs are located in a slightly deeper plane,
approximately 30µm in diameter with relatively
homogeneous brightness
(Messmer et al., 2006)
Presumed GCs are large to giant hypo-reflective
round to oval-shaped with a nucleus displaced
peripherally, sometimes crowded in groups and
visible throughout the epithelium
(Zhivov et al., 2006)
Presumed GCs are relatively larger cells about 25µm
(Rath et al., 2006)
Presumed GCs can be easily recognized by their size,
highly reflective pixels depict cell walls or wide
intercellular spaces with high contrast
(Messmer, 2008)
Large to giant hyper-reflective round to oval-shaped
cells with a nucleus displaced peripherally sometimes
crowded in groups
Image Sample Size for Goblet Cell Density Determination 81
(Efron et al., 2009)
Presumed GCs are slightly larger and paler than
surrounding non-secretory epithelial cells
(Villani et al., 2011a)
Presumed GCs are roundish, slightly larger and
brighter than surrounding cells of approximately
30µm
(Wei et al., 2011)
Presumed GCs have hypo-reflective nucleus or with
relatively homogeneous brightness, larger than
surrounding epithelial cells, crowded in groups or
dispersed in the layer of intermediate epithelial cells
(Mastropasqua et al., 2013)
Presumed GCs appeared large, hyper-reflective and
oval-shaped with hypo-reflective nuclei, larger than
the surrounding epithelial cells, crowded in groups or
dispersed in the layer of epithelial cells
No studies to date have identified the optimal sampling of images captured with
LSCM when evaluating GCD. Therefore, using the identification strategies described
above, this study aimed to determine the acceptable level of precision for GCD
calculation.
Image capture
Conjunctival LSCM was performed using the Heidelberg Retinal Tomograph (HRT
III) equipped with a Rostock Corneal Module (Heidelberg Engineering GmbH,
Heidelberg, Germany). One eye (the eye preferred by the participant) was examined.
The eye was anaesthetized with 0.4% oxybuprocaine hydrochloride (Chauvin
Image Sample Size for Goblet Cell Density Determination 82
Pharmaceuticals Ltd, UK). To optimise the quality of CIC specimens collected
following LSCM, no drop of ocular gel was used between the ocular surface and the
front of the TomoCap (diameter 12 mm). The participant was instructed to direct their
gaze towards the opposite direction to the region of measurement (nasal bulbar
conjunctiva). The centre of the surface of the TomoCap was positioned on the
conjunctiva about 2 to 4 mm from the limbus. The appropriate layers of the
conjunctiva were observed and the focal plane was gradually moved into the
conjunctival epithelial tissue between 10 to 44 µm (Zhang et al., 2011) until the GC
groups were visualised. The GC groups of the nasal bulbar conjunctiva were scanned
while moving the applanating lens in X, Y and Z path at nine different locations
(approximating a 3 x 3 grid). A sequence of 30 image frames was captured from at
least three different depths.
Target fixation protocol
A fixation protocol was developed in order to assure that the same area of the nasal
bulbar conjunctiva was examined at each visit. The external fixation light of the
HRTIII was set at a distance of 30 cm and separated horizontally from the eye of the
participant by 15cm, creating an approximately 60° angle between the centre of
TomoCap and the centre of the pupil. This set up of the target allowed the imaging of
the nasal bulbar conjunctiva at approximately 2 to 4 mm from the limbus.
Image selection criteria
High quality images suitable for cell count containing abundant GCs (Doughty,
2012b) and included GCs identified according to their size, shape and reflectivity, i.e.
25-30 µm in diameter (Kobayashi et al., 2005; Zhivov et al., 2006), hyper-reflective,
(Messmer, 2008) bigger than surrounding cells (Villani et al., 2011a), round
(Messmer, 2008) to oval (Pisella et al., 2001) in shape and sometimes with a visible
nucleus (Hong et al., 2010) were sequentially selected such that they were non -
overlapping by more than 20%.
Determination of goblet cell density
Quantification of cells was conducted using the manual cell count mode of the
Heidelberg Eye Explorer software (Heidelberg Engineering GmbH, Heidelberg,
Image Sample Size for Goblet Cell Density Determination 83
Germany). The manual cell count mode has an integrated tool to manually count cells
by clicking on each cell of the image selected. A blue marker indicates that the cell is
counted and the cell density (number of cells per mm²) is displayed in the cell count
results window. The L-form rule was used to count GCs from the 400 x 400 µm²
images; that is, all GCs that lie on the left and the lower border were counted. None of
the GCs on the upper and the right border were counted.
Validation for image sampling approach
A sequence of approximately 30 image frames was captured from 10 symptomatic CL
wearers. The reason symptomatic CL wearers was a selection criteria in this sampling
study was that the variation of the standard deviation (SD) of 3 previous images was
higher than for those not wearing CLs. The variance (every possible combination) of
3 to 30 images of GCs was plotted against the number of images taken, to determine
the point at which variability became relatively constant. This analysis revealed that a
minimum of 11 images not overlapping by more than 20% were necessary to
determine the average of GCD at each examination in order to ensure an acceptable
level of precision for the GCD calculation in the main longitudinal study. This
approach resulted in the mean GCD of 485 and a SD ± 40 cells/mm² (Figure 7).
Image Sample Size for Goblet Cell Density Determination 84
Figure 7.1. Scatterplot of standard deviation vs. the number of images for 10
symptomatic CL wearers. Image sampling analysis for LSCM. Each data point
represents the standard deviation of 3, 4, 5... and so on up to 30 images plotted
against the number of images taken from 10 symptomatic contact lens wearers. A
minimum of 11 random images were necessary to determine the average of GCD.
The average standard deviation was approximately ± 40 cells/mm².
7.3.2 Conjunctival Impression Cytology
Sample collection
A few minutes after performing LSCM, the same eye was anaesthetized again and the
centre of a Biopore membrane (Millicell cell culture inserts; Millipore Corp, Cork,
Ireland, United Kingdom) was gently applied to the nasal bulbar conjunctival surface
at approximately 2 to 4 mm from the limbus. The sample was allowed to air dry and
then immersed in 95% methanol for fixation using a well culture plate sample holder.
The sample was then refrigerated under 4 ºC for no more than 24 hours.
To verify the location of the impression and the integrity of the exposed bulbar
conjunctiva (Figure 7.2), a slit lamp examination with fluorescein was conducted
under cobalt blue illumination with a yellow filter (a close match to a Wratten #12
Yellow). After any minimally invasive technique that can potentially disrupt
0
10
20
30
40
50
60
70
80
90
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
SD
GC
D C
ells
/mm
²
Number of images
Participant 1 Participant 2 Participant 3 Participant 4 Participant 5
Participant 6 Participant 7 Participant 8 Participant 9 Participant 10
Image Sample Size for Goblet Cell Density Determination 85
superficial cell from the ocular surface, it is prudent to evaluate if the techniques used
to assess the cells affect the ocular health of the participants. As this procedure is
performed using fluorescein stain, a positive staining is observed after CIC, along
with the use of fixation target, indicating that superficial cells are indeed removed
approximately 2 to 4 mm from the limbus.
Target fixation protocol
In order to assure that the same area of the nasal bulbar conjunctiva was examined at
each visit by both techniques, a fixation protocol for CIC was developed. A fixation
target was set at a distance of 30 cm and separated horizontally from the eye of the
participant by 15cm, creating an approximately 60° angle between the centre of
Millicell cell culture inserts and the centre of the pupil. This set up of the target
allowed the sample collection of the nasal bulbar conjunctiva at approximately 2 to 4
mm from the limbus of the same eye examine previously with LSCM.
Figure 7.2. Photograph of the eye under examination using slit lamp and fluorescein
conducted under cobalt blue illumination with a yellow filter. Location of the
impression and evaluation of the integrity of the exposed bulbar conjunctiva is
observed.
Image Sample Size for Goblet Cell Density Determination 86
Sample staining procedure
The staining procedure was performed using Giemsa stain according to the following
guidelines from the manufacturer (Sigma-Aldrich): Millicell inserts with more than
60% of cellular material across the field of the filter were assessed. The same well
culture plate sample holder was used to retain the specimens during staining. The
specimen was allowed to air dry at room temperature, the Giemsa stain was diluted
1:20 with deionized water, and the specimen was immersed in the diluted Giemsa
solution for 30 minutes (Figure 7.3). The sample was rinsed with deionized water and
the filter was carefully detached from the plastic ring and prepared in glass slides
prior to microscope examination.
Figure 7.3. Staining procedure using Giemsa stain. (A) More than 60% of the filter is
covered in cell material. (B) Staining procedure in a well plate sample holder. (C) The
filter of the Millicell cell culture insert detached from the plastic ring and ready for
imaging onto a slide and covered with a coverslip.
Image capture
Images of the sample were captured by a Leica DM2500 microscope (Leica
Microsystems) to visualize the specimen collected; this system had a magnification of
x200 and field of view of 640 x 480 µm². The slide holder of the microscope was
moved in X and Y directions along the sample with the aim of collecting images non-
overlapped by more than 20%. Approximately 10 images with non-disrupted cell
material and usually with over layered epithelial cells were taken from each sample.
Morphological identification of GCs using Giemsa stain was undertaken according to
the image selection criteria described below.
A B C
Image Sample Size for Goblet Cell Density Determination 87
Image selection criteria
Images of the sample with no-disrupted cell material that contained GCs
approximately 25 to 30µm in diameter were selected. The cells had a pale membrane
with defined borders, and a visible nucleus localised centrally, although sometimes
eccentrically in bigger cells (approximately 30 µm) (Doughty, 2011a). GCs were
easily differentiated from surrounding cells because of their balloon-like appearance
and cell size (Figure 6.1).
Validation of image analysis approach for CIC
A similar analytical approach performed previously for assessing GC with LSCM was
used to evaluate CIC sampling. The mean GCD for each specimen was determined by
averaging every combination of 3 to 10 images of non-disrupted cell material. This
analysis revealed that a minimum of 5 images were necessary to determine the
average of GCD at each examination in order to ensure an acceptable level of
precision for the GCD calculation in the main longitudinal study. This approach
resulted in the mean GCD of 355 ± 152 cells/mm² (Figure 7.4).
Image Sample Size for Goblet Cell Density Determination 88
Figure 7.4. Scatterplot of standard deviation vs. the number of images for 10
symptomatic CL wearers. Image sampling analysis for CIC. Each data point
represents the standard deviation of 3, 4, 5... and so on up to 10 images plotted the
against number of images taken from 10 symptomatic contact lens wearers. A
minimum of 5 images was necessary to determine the average of GCD. The average
of the standard deviation was approximately ± 152 cells/mm²
Determination of goblet cell density
Quantification of cells was conducted using the cell counter plugin of ImageJ
software (National Institutes of Health, Bethesda, Maryland). The manual cell count
software has an integrated tool to facilitate manual counting of cells by clicking on
each cell of the image selected. Each click marks the cell with a coloured cross and
adds the cell number to a tally sheet. The total cell count is displayed in the cell count
results window. The L-form rule explained in Section 7.1.4 ‘Determination of goblet
cell density’ was also used to count GCs from the 640 x 480 µm² images at 200x
magnification. The images were acquired through a video camera attached to the
microscope. Using the scale bar, the total area of the image is known and so the GCD
can be calculated by the number of GCs per unit area in cells/mm² using the following
formula: number of GCs per image / area mm².
50
70
90
110
130
150
170
190
210
230
0 1 2 3 4 5 6 7 8 9 10
SD
GC
D c
ells
/mm
²
Number of images
Participant 1 Participant 2 Participant 3 Participant 4 Participant 5
Participant 6 Participant 7 Participant 8 Participant 9 Participant 10
Image Sample Size for Goblet Cell Density Determination 89
7.4 DISCUSSION
The results of this analysis ensured an acceptable level of precision for GCD
calculation using CIC and LSCM, and also minimized the variability in the final
results of this longitudinal investigation. Based on the results obtained in this analysis,
the average of five and 11 images were necessary to determine GCD using CIC and
LSCM, respectively. Therefore, at each visit, this number of images was collected per
participant.
General Methodology and Research Plan 91
General Methodology and Chapter 8:
Research Plan
8.1 STUDY DESIGN
This study was a prospective longitudinal observational case-controlled clinical study.
GCD was measured in individuals who met the inclusion/exclusion criteria before and
during 6 months of CLW. GCD was measured using both LSCM and CIC in number
of cells/mm². GCD in individuals who developed DE from CLW was compared to
those who did not develop DE symptoms and to healthy controls.
The testing procedures and analysis of images from both techniques was conducted in
a masked fashion. Ethical clearance was provided by the University Human Research
Ethics Committee, and written informed consent was obtained from all participants
(Appendix 1).
8.2 RECRUITMENT
Staff and students at QUT were approached to participate in the study via email. The
participants who were fitted with CLs had contact lenses and ocular care provided
free of cost for the duration of the study (6 months). Participants were offered the
benefit of a new option to correct their vision, i.e. the possibility to use CLs instead of
spectacles or the use of both, according to participant needs. The group that did not
wear CLs (the control group) was reimbursed for travel expenses and out-of-pocket
expenses with $80 Coles–Myer vouchers at the end of the study.
Participants made four visits to the Vision Test Room Q531, located on level 5 at
IHBI. The baseline examination was followed by appointments at 1 week, 1 month,
and 6 months after baseline. There was a CL fitting visit for the CL wearing group at
least 24 hours after the baseline appointment.
8.3 SAMPLE SIZE CALCULATION
The number of participants that were enrolled into this study was determined by a
sample size calculation based on two previous studies (Simon et al., 2002 and
General Methodology and Research Plan 92
Doughty, 2012) with a similar design to the present study. The mean GCD of healthy
non-CL wearers of 620 ± 154 cells/mm² was reported using a light microscope
medium power of approximately 200x magnification (Doughty, 2012b). Another
group reported the reduction of GCD after 6 months of CLW to be 26.7% compared
to the control group (non-CL wearers) (Simon et al., 2002). These data were used to
calculate the minimum sample size of 23 participants per group, allowing of 20%
attrition. This analysis gave 90% power, with a type I error of 5% to detect the
difference in GCD between the two groups if one exists. A total of 25 participants per
group were enrolled to further account for potential attrition of participants for the
follow-up visits
8.4 STUDY POPULATION
Following written informed consent, 110 participants were screened and after
applying the inclusion/exclusion criteria (Sections 8.4.1 ‘Inclusion criteria’ and 8.4.2
‘Exclusion criteria’), 83 healthy participants from multiple ethnicities were enrolled in
the study (Figure 8.1). The main reason for exclusion was failed DEQ-5 (5
participants), followed by participants taking medication that is known to affect the
tear film such as antidepressants (3 participants), and also personal decision (3
participants). Other reasons for the exclusion of a further 6 potential participants were
systemic conditions, ocular surgery and trauma, corneal scarring, pregnancy and
ocular allergy. At baseline, another 10 participants could not be fitted due to CL
fitting intolerance (5 participants), failure to attend the visits (4 participants) and
recurring conjunctivitis (1 participant). Sixty (60) participants were fitted into a
conventional hydrogel lens: (ocufilcon D) Biomedics 1 day Extra by CooperVision
with 55% water content and 27 Dk/t for daily replacement basics. Twenty-three (23)
age-balanced non-contact lens wearers served as the control group. Approximately
40% and 30% of the participants enrolled were Caucasians and Saudi respectively,
and the remainder of the population were of other ethnicities as shown in Figure 8.1.
After 1 week of CLW, participants completed a DE questionnaire specially designed
for soft CL wearers (recommended by DEWS 2007), i.e. the CLDEQ-8. This
questionnaire was developed and validated to examine the distribution of DE
symptoms in soft CL wearers and consists of 8 questions. Potential scores range from
0 to 37 points. Scores of 17 or more out of 37 represents DE. Twenty-five (25)
participants who reported scores of 17 or more points were assigned to the group
General Methodology and Research Plan 93
entitled ‘symptomatic of CL-induced DE’. The participants who reported scores of 16
or below points were considered ‘asymptomatic of CL-induced DE’ (Figure 8.2).
Figure 8.1. The bar chart represents the country of origin in percentage of the study
population that was examined at baseline (N = 83).
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
2.4
2.4
2.4
3.6
3.6
3.6
28.9
39.8
0 5 10 15 20 25 30 35 40 45
Bhutan
Canada
Chile
Jordan
Nigeria
Philippines
Russia
Singapore
South Africa
South Korea
United Kingdom
Malaysia
New Zeland
Vietnam
China
Colombia
India
Saudi
Australia
Percent
Co
untr
y o
f o
rigin
General Methodology and Research Plan 94
Figure 8.2. A schematic representation of the study population enrolled, excluded, the
number of examinations at baseline and those assigned to DE groups after the 1-week
visit.
Flow chard of the study population
Total participants screened = 110
Total participants enrolled = 93
Total participants at baseline = 83
Participants
excluded = 17
Total participants not
assigned to a group = 10
Failed DEQ-5 = 5
Medication use = 3
Personal decision = 3
Other reasons = 6
CL intolerance = 5
Lost to follow-up = 4
Recurring conjunctivitis = 1
Non-CL wearers = 23
(age-balanced controls)
Participants fitted
with CL = 60
1 week of CLW
CLDEQ-8
Assessment
17 points or more 16 points or less
Symptomatic of CL-
induced DE = 25
Asymptomatic of CL-
induced DE = 35
General Methodology and Research Plan 95
8.4.1 Inclusion Criteria
Aged 18 to 50 years old
Signed written informed consent
Able and willing to participate in the study
8.4.2 Exclusion Criteria
A history of CLW in the last 6 months
Current pregnancy or breastfeeding
History of ocular trauma or surgery
Ocular surface dysfunction
Classification as symptomatic for DE based on answers to the DE
questionnaire (DEQ-5) (Chalmers et al., 2010)
Current or long-term use of topical ocular medication
Ocular disease or systemic disease that may affect the conjunctiva
An additional exclusion criterion was applied to the CL wearers based on their
refractive error. Participants were not included if they had astigmatism of 1.50 D or
more, myopia of -7.00 D or more and hyperopia of +2.00D or more. This exclusion
criteria was applied because of the lens design used on this study (Biomedics 1 day
Extra).
8.5 SCREENING AND BASELINE
All participants screened in the study underwent a detailed recording of ocular and
health history. The history-taking aimed to obtain information about ethnicity, the use
of medication, current pregnancy and/or breastfeeding, menstrual cycle and general
ocular history such as past trauma, surgery and CLW. After the participants were
determined suitable for the study per the exclusion/inclusion criteria, they were asked
to complete the DEQ-5. Provided a score of 7 or less points was achieved; the
individual underwent ocular screening procedures including visual acuity and
subjective refraction, followed by ocular surface and DE tests. All examinations were
performed in the morning between 7:00 am to 12:00 noon to avoid potential influence
of diurnal variations on the measures.
General Methodology and Research Plan 96
After determination of eligibility, each participant was assigned to wear disposables
CL in daily basis or the control group (this was, with one exception, based on
refractive error requirements).
At each visit participant DE and GCD measurements were taken.
DE and ocular surface assessments were as follow:
CLDEQ-8 assessment for CL wearers and DEQ-5 for controls
Non- invasive break- up time (NIBUT)
Ocular surface staining (OSS)
Phenol red thread (PRT) test
GCD assessment as described in Chapter 7 Sections 7.1 ‘Conjunctival laser scanning
confocal microscopy’ and 7.2 ‘Conjunctival impression cytology’:
Nasal bulbar conjunctiva imaged using LSCM (HRTIII)
CIC conducted at same location evaluated with LSCM
8.6 DRY EYE AND OCULAR SURFACE ASSESSMENT
The criteria from the DEWS was applied to evaluate the tear film and diagnose DE.
Each participant completed a DE questionnaire (DEQ-5) at baseline and underwent
NIBUT testing, recorded in seconds. Fluorescein staining and the Efron grading
system was used to evaluate the ocular surface and graded on a 0 - 4 scale. Tear
volume was assessed using the PRT test, recorded in millimetres. Each test, as
described below, had a cut-point indicative of DE and designated as a ‘fail’ to be
included in the study, participants had to ‘pass’ the DEQ-5 and at least one of the
three DE clinical tests to be eligible for the study.
8.6.1 DEQ-5 Questionnaire
The DEQ-5 Questionnaire was selected out of eight recommended DE symptom
questionnaires evaluated by the DEWS such as The Women’s Health Study
(Schaumberg et al., 2003); International Sjögren Classification (Vitali et al., 2002);
General Methodology and Research Plan 97
Schein (Schein et al., 1999); McMonnies (McMonnies, 1986); OSDI (Schiffman et
al., 2000); CANDEES (Doughty et al., 1997); DEQ (Begley et al., 2003) and IDEEL
(Rajagopalan et al., 2005). The DEQ-5 questionnaire was selected because it detects
mild to moderate DE symptoms that could potentially develop in CLW. This unique
questionnaire is the shorter, validated version of DEQ and measures DE symptoms in
terms of frequency and intensity and only requires 1 to 2 minutes for completion. The
self-assessment questionnaire rates the severity of DE symptoms on a scale that
ranges from ‘I don’t have dry eye’ (0) to ‘extremely severe dry eye’ (5). The
questionnaire is composed of five questions related to the most relevant DE
symptoms. For a participant to be considered as asymptomatic, the total score should
be 6 or less points (Chalmers et al., 2010).
8.6.2 CLDEQ – 8 Questionnaire
The CLDEQ-8 questionnaire was designed for CL wearers. This questionnaire was
originally designed in parallel with the DEQ-5 for use in patients who do not wear
CLs (Chalmers et al., 2012a). CLDEQ-8 was developed to assess DE symptoms
among CL wearers specifically. Soft CL wearers report a different pattern of
symptoms compared with those who do not wear soft CLs (Chalmers et al., 2009).
This validated questionnaire consists of 8 questions and provides a score ranging from
of 0 to 37, in which DE symptoms represents scores of 17 or more out of 37 points.
8.6.3 Non-Invasive Break- Up Time Test
The DEWS recommends NIBUT testing for tear stability assessment when the results
of a study are potentially susceptible to selection bias in research trials. NIBUT
measurements provide moderately high sensitivity (83%) with good overall accuracy
(85%) (Smith, 2007). The test involved the use of a keratometer (KM-1 Takagi Seiko
co., Ltd, Japan) in combination with a fine grid insert. The time between full opening
of the eyelids after a complete blink and the first observed break in the tear film was
measured using a digital timing device. The median of three readings was recorded.
Measurements from the two eyes for each measure were averaged to give a single
value for each participant. An average reading of 11 seconds or less indicated poor
NIBUT and was recorded as a ‘fail’.
General Methodology and Research Plan 98
8.6.4 Ocular Surface Staining
Numerous tools can be used to evaluate the ocular surface. Fluorescein, lissamine
green and rose bengal are some of the dyes used most commonly during
biomicroscopy to assess the ocular surface integrity. Disruption of the stain suggests
damage or irregularity of the ocular surface epithelium. In this study, the Efron
Grading Scale System was used to score ocular surface integrity on a 0 - 4 scale.
To stain the ocular surface, a drop of saline was placed on a fluorescein-impregnated
strip, which was then gently touched against the tarsal conjunctiva. Blue-light slit-
lamp biomicroscopy and a Boston yellow filter were used to evaluate corneal and
conjunctival staining, conjunctival redness and papillary conjunctivitis. The degree of
staining was graded from 0 to 4 (normal to severe staining). Moderate (grade 3) to
severe (grade 4) was recorded as a sign of DE. Measurements from the two eyes for
each measure were averaged to give a single value for each participant (Figure 8.3).
0 – normal 1 – trace 2 – mild 3 – moderate 4 – severe
Figure 8.3. Efron grading scale system for ocular surface assessment.
8.6.5 Phenol Red Thread Test
The PRT test is a minimally invasive test of tear quality and utilises a thread treated
with the pH indicator phenol red (phenolsulfonphthalein).The repeatability of this test
has been studied extensively. Repeated measure analysis of variance revealed no
significant difference in the PRT values obtained on different days (Blades and Patel,
1996; Cho, 1993; Little and Bruce, 1994; Tomlinson et al., 2001). A PRT test (Tianjin
Jingming New Technological Development Co., Ltd, China) was placed in the lower
conjunctival sac on the temporal side of each eye for 20 seconds without anaesthetic
with both eyes open and the length of thread that turned yellow was measured against
the scale on the test package. Measurements from the two eyes for each measure were
General Methodology and Research Plan 99
averaged to give a single value for each participant. A wet length of approximately 10
mm or less was considered as a ‘fail’.
Table 8.1. DE and ocular surface assessment cut-point indicative of CL-induced DE
symptoms
Test Asymptomatic of CL-induced DE Symptomatic of CL-induced DE
CLDEQ-8 (0 – 37) 0 - 16 17 - 37
NIBUT (s) 1 - 10 ≥ 11
OSS (0 - 4) 0 - 2 3 - 4
PRT (mm/20s) ≥ 11 ≤ 10
CLDEQ-8, Contact Lens Dry Eye Questionnaire- 8; NIBUT, non-invasive tear break-up time;
OSS, ocular surface staining; PRT, phenol red thread test
8.6.6 Dry Eye and Ocular Surface Criteria for Enrolment
Various studies have shown disagreement between DE symptomatology and the
results of corresponding clinical tests (Bjerrum, 1996; Hay et al., 1998; Schein et al.,
1997), with only 57% of symptomatic subjects presenting clinical signs of DE (Hay et
al., 1998; Pflugfelder et al., 1998; Schein et al., 1997). This finding has been
attributed to the aetiology and pathophysiology of DE (McCarty et al., 1998). As a
result, a single objective test without subjective symptoms is not sufficient for a
diagnosis of DE (Report of the International Dry Eye WorkShop (DEWS), 2007).
At baseline, all participants met the DE criteria suggested by DEWS in 2007 in order
to be enrolled in the study. Participants at baseline passed the DEQ-5 questionnaire as
well as at least one of the three DE clinical tests previously described. All
examinations were performed in the morning by the same examiner.
8.7 CONTACT LENS FITTING AND FOLLOW-UP VISITS
CL fitting took place at least 24 hours after the baseline examination, due to the use of
ocular topical anaesthetic and the potential mild debridement of the nasal conjunctiva
associated with CIC and LSCM. Each member of the CL group was asked not to use
the CL for 24 hours after each visit.
These examinations were conducted at 1 week ± 3 days and after 1 and 6 months ± 7
days of CLW; all groups had the same follow-up schedule. The aim of these
General Methodology and Research Plan 100
examinations was to ensure participants were safe to wear CLs and to collect
longitudinal data to address the research questions regarding GCD assessed by both
techniques.
8.8 MASKING AND RANDOMIZATION
The files and images taken were labelled by observer one (YA) with a unique
identification code for each participant at each visit. No names or dates of birth were
used in order to maintain anonymity and masking conditions. A separated data set
with the information linked to the personal information of the participant was saved in
a QUT mainframe drive with limited access.
Cell counts were performed by observer two (LH). Approximately 30 frames from
LSCM and 10 frames from CIC were obtained and 11 and 5 images with the selected
criteria were randomized for cell count using LSCM and CIC, respectively.
8.9 RATIONALE
8.9.1 Rationale for 6-Month Study
A study duration of 6 months was chosen for a number of reasons. A key aspect of
the longitudinal phase of this project was to differentiate between asymptomatic
versus DE symptomatic CL wearers. Any differences in physiological response
between these two sub-groups may be subtle and may take some time to develop. A 6-
month time frame struck a balance between a sufficient time to observe significant
changes and completing a PhD within the prescribed time frame. In addition, the
literature describing adaptive changes to CLW, both in terms of subjective responses
and observed objective signs of physiological change, typically run from 3 to 6
months, therefore a 6 month study dataset that can be compared to the existing
literature in terms of changes observed during the initial phase of CLW (Cavanagh et
al., 2002; du Toit et al., 2001b). Finally, as referred in 'individual contributions'
section in this document, this project is aligned with a study by Yahya Alzahrani, who
also required a 6-month observation time.
General Methodology and Research Plan 101
8.9.2 Rationale for Using Conventional Hydrogels
This material was selected for this study (rather than silicone hydrogels daily
disposables) due to anticipated physiological effect on the eye. Silicone hydrogel
lenses are known to have a very low physiological effect on the eye. The difference in
the physiological impact of silicone hydrogels and conventional hydrogels for daily
wear are not only in their material and water content but also reports suggest that
silicone hydrogels provide increased comfortable wearing time, with reduced
symptoms of dryness at the end of the day (French and Jones, 2008). On the other
hand, the performance of conventional hydrogels is limited to the water content,
which indicates that the only way for the cornea to receive an adequate percentage of
oxygen is through adequate water content (Tighe, 2006). The oxygen transmitted to
the cornea during silicone hydrogel lens wear is primarily through the silicone
material of the lens, rather than the water content, therefore, conventional hydrogels
afford lower oxygen transmissibility than silicone hydrogels (Efron et al., 2007). In
fact, there are a number of studies which have demonstrated that corneal swelling is
reduced with silicone hydrogels, even after as little as one hour dozing during daily
lens wear, (Hamano et al., 2008) and the event of hypoxic complications such as
striae, folds, blebs and microcysts are also reduced (Brennan et al., 2008; Fonn et al.,
2005; Keay et al., 2001; Stretton et al., 2003). Recent studies have also demonstrated
that daily wear of silicone hydrogels reduce limbal hyperaemia compared to
conventional hydrogels (du Toit et al., 2001a; Riley et al., 2006).
In conclusion, lenses that contain silicone are more likely to preserve the physiology
of the ocular surface in daily wear. Therefore, it has been assumed in this
investigation that conventional hydrogels will potentially affect conjunctival GCD in
a shorter timeframe than silicone hydrogels.
Association between LSCM and CIC for Goblet Cell Density Assessment 103
Association between LSCM and Chapter 9:
CIC for Goblet Cell Density
Assessment
9.1 PREFACE
Ex vivo cell analysis obtained from CIC has been used extensively in the past 30
years to report conjunctival GCD while LSCM examination is a less- known
application to explore and report cell densities from the ocular surface. However, few
reports have been published on the assessment of GCD using LSCM in ocular surface
disorders and healthy individuals. Reports vary regarding the GCD measured using
LSCM of the four cardinal points of the bulbar conjunctiva (nasal, superior, temporal
and inferior) in healthy individuals as well as ocular surface disease or damage.
GCD using LSCM in health individuals at the four cardinal points ranged from an
average of 111 ± 58 cells /mm² (Efron et al., 2009) and 432 ± 72 cells/mm² (Zhu et
al., 2009). One additional report indicated an average of 260 cells/mm² in the nasal
bulbar conjunctiva (Efron et al., 2010a). The comparison of LSCM with CIC has been
investigated in participants with ocular surface disorders. The lower bulbar
conjuctival GCD has been noted to be greater with LSCM than CIC in Sjögren DE
syndrome (332 ± 137 vs 200 ± 141cells/mm²) (Hong et al., 2010) yet similar in the
case of chemical burns (136 ± 79 vs. 121 ± 66 cells/mm2) (Le et al., 2010).
Impression cytology is considered the ‘gold standard’ technique for assessing cell
morphology of the ocular surface, however, is more invasive than assessment using
LSCM. A scale system has been developed by Nelson and co-workers (Nelson, 1988)
that reflects metaplastic changes to epithelial cells as well as changes in the number of
GCs using CIC. This scale has been used to identify cells on the ocular surface in eyes
treated with both preserved and preservative-free glaucoma therapies (Ciancaglini et
al., 2008; Mastropasqua et al., 2013) for both CIC and LSCM techniques. In two
disease models, a positive correlation of GCD using LSCM and CIC was
demonstrated in people with Sjögren syndrome (ρ = 0.908; p < 0.05) (Hong et al.,
2010) and chemical burns (ρ = 0.946; p = 0.000) (Le et al., 2010). The CIC technique
Association between LSCM and CIC for Goblet Cell Density Assessment 104
has been widely used for the past three decades to report GCD; however, limitations
of this technique, mainly relating to its invasive nature, have been raised previously
by many authors. There is therefore a need to understand the utility of the less-
invasive LSCM compared to the current standard of GC assessment, namely CIC.
This study reports, for the first time, the association between the gold standard CIC
technique and the new, non-invasive technique of LSCM in a healthy population.
Demonstration of an association between these two techniques will serve to validate
LSCM as a viable alternative procedure to assessing GCD in human populations. The
hypothesis is that there will be a strong association between the two measures and that
LSCM can replace CIC as a less-invasive measure of GCD in the conjunctiva.
9.2 PURPOSE
To determine the association between conjunctival GCD assessed using in vivo
LSCM and ex vivo CIC.
9.3 METHODS
This was a cross-sectional study of conjunctival GCD measured using in vivo LSCM
and ex vivo CIC. A total of 90 participants (44 women, 46 men; age 30.8 ± 8.5 years)
were enrolled in the study after meeting inclusion/exclusion criteria described in
Chapter 8 Sections 8.4.1 ‘Inclusion criteria’ and 8.4.2 ‘Exclusion criteria’. All
participants completed the DEQ-5 questionnaire and underwent an ocular surface and
DE examination. The DE and ocular surface criteria described in Section 8.6.6 ‘Dry
eye and ocular surface assessment’ was also applied for enrolment. NIBUT was
recorded as described in Section 8.6.3 ‘Non-invasive break-up time test’. The degree
of ocular surface staining with fluorescein was graded from 0 to 4 according to the
validated Efron grading scale system and a PRT (Tianjin Jingming New
Technological Development Co., Ltd, China) was placed in the lower conjunctival sac
and recorded as described in Section 8.6.5 ‘Ocular surface staining’ of Chapter 8.
All examinations were performed in the morning by the same examiner. Since the GC
distribution is apparently random using LSCM throughout the bulbar conjunctival
tissue, we assume that the nasal bulbar at approximately 2 to 4 mm from limbus area
Association between LSCM and CIC for Goblet Cell Density Assessment 105
is a representative and reliable approach that roughly correspond to the CIC
technique. Hence, we adopted the sampling approach described in Chapter 7.
9.4 STATISTICAL ANALYSIS
The association between CIC and LSCM was assessed using Spearman correlation
and bootstrapped confidence intervals (95%). To analyse the agreement between
measurements on the same participant, a regression approach for non-uniform
differences was carried out using Bland-Altman technique (Bland and Altman, 1999)
with linear regression and 95% LoA applied (Efron et al., 2010b). Global values of
GCD were used for this analysis (the average of 5 and 11 images for CIC and LSCM,
respectively). SPSS for Windows Version 16 (SPSS Sciences, Chicago, IL) was used
for this statistical analyses.
9.5 RESULTS
The Spearman’s rho correlation revealed a strong positive relationship between GCD
assessed with CIC and LSCM (ρ = 0.66, 95% CI: 0.52 - 0.77). A significant
difference in GCD was found when assessed using CIC and LSCM (466 ± 51
cells/mm² and 475 ± 41 cells/mm², respectively; paired t = 2.26, p = 0.026); the mean
difference between the two measurements was 9 cells/mm², which in the context of
the variability in cell densities is considered equivalent.
A Bland-Altman plot of GCD obtained using the two methods is shown in Figure 9.1.
This plot shows the relation between differences in GCD vs. mean GCD. On the
graph, the middle line represents the linear regression. The upper and lower lines
represent the 95% limits of agreement. Regression analysis revealed an R2 of 0.49 (p
< 0.001). The downward slope of the regression line indicates that, for higher mean
CGD values, a higher value was assigned to GCD as assessed with LSCM and a
reduced spread of data is associated with lower GCD values obtained with CIC.
The results of our assessment of ocular surface integrity and dry eye assessment of the
participants in this experiment are summarised in Table 9.1. The 90 participants were
asymptomatic for dry eye based on results of the DEQ-5 (scores of < 7 points) and all
the participants passed the ocular surface staining with scores of < 2 points using the
Association between LSCM and CIC for Goblet Cell Density Assessment 106
validate Efron grading scale. Only 6 participants failed the PRT test with scores of
<10 mm/20 s and 23 participants failed NIBUT with scores of >10 s of tear break.
Table 9.1. Ocular surface integrity and dry eye assessment of participants
Statistics DEQ-5 (0-17) NIBUT (s) OSS (0 - 4) PRT (mm/20s)
Mean ± SD 3 ± 2 13 ± 6 0 ± 1 20 ± 8
Min - Max 0 - 8 4 - 30 0 - 2 6 - 40
DEQ-5, 5-Item Dry Eye Questionnaire; NIBUT, non-invasive tear break-up time; OSS,
ocular surface staining; PRT, phenol red thread test
Figure 9.1. Bland-Altman plot of the differences in GCD between the two techniques versus
the average. The middle, heavier line represents the linear regression and the lighter upper
and lower lines represent the 95% LoA. Each of 90 data points represents the average value
of 5 and 11 images with CIC and LSCM, respectively. LSCM and CIC agree at any time
between 48.52 to 213.31 (upper bound) and -0.43 to -0.08 (lower bound).
-150
-100
-50
0
50
100
150
350 400 450 500 550 600 650
Dif
fere
nce
(L
SC
M-C
IC)
cell
s/m
m²
Average (cells/mm²)
Association between LSCM and CIC for Goblet Cell Density Assessment 107
9.6 DISCUSSION
This study reports a strong association between CIC and LSCM for the assessment of
GCD in healthy participants. This finding is consistent with previous reports that
examined the correlation of GCD measurements using these techniques. These studies
from the literature have positively correlated GCD measurements assessed with CIC
and LSCM in patients with pathology, such as chemical burns on the conjunctiva (r =
0.929) (Le et al., 2010) and Hong, et al. (2010) ρ = 0.908 in Sjögren syndrome
patients. The higher reported correlations between CIC and LSCM in diseased eyes in
the Le and Hong studies is possibly due the truncated sample and sampling bias
(Hong et al., 2010).
Similarly to these previous results, readings from CIC were slightly lower than those
made from LSCM (p = 0.026), as shown in Figure 9.1. This phenomenon could be
attributed to (a) the improbability of all cells in the sample region attaching to the
filter at the time of peeling from bulbar conjunctiva when performing CIC, and (b) the
inability of CIC to sample cells at deeper layers of the conjunctival epithelium, unlike
LSCM which can scan cells at different depths of the epithelium.
The distribution of the GCD values in the Bland–Altman plot indicate that the higher
the GCD average, the greater the difference between GCD values obtained with
LSCM and CIC. The reason for this difference profile is unclear.
Systematic errors related to sampling techniques are the source of variations between
invasive and non-invasive techniques. For example, staining methods can vary using
the CIC technique according to the filter used to collect the cells. Conventional
cellulose acetate filters allow the observation of cells under a light microscope using
coloured stains. For immunofluorescence staining, however, the filter must have
specific properties such as mixed cellulose esters and larger pore size. A few reports
in the literature have mentioned that different filter types can improve sample
consistency and cell attachment (Albietz, 2001; Doughty, 2012b). However, in some
studies using conventional cellulose acetate filters, greater pressure was applied to the
conjunctiva for longer periods of time during sample collection in order to obtain the
same outcomes as those obtained with mixed cellulose esters.
Association between LSCM and CIC for Goblet Cell Density Assessment 108
There appears to have been no validated approach to the number of images acquired
in previous studies when attempting to correlate GCD assessed using LSCM and CIC.
One study used an average of 3 images from each of the cardinal points of the bulbar
conjunctiva (nasal, superior, temporal and inferior) using LSCM. The same study
used an average of 3 consecutive images from only two sites of the interpalpebral
conjunctiva (nasal and temporal) when performing CIC (Le et al., 2010). Another
study using LSCM captured images from the superior bulbar conjunctiva in the Z-axis
and averaged 4 images for the total GCD. These measurements were correlated with
an average of 3 consecutive images obtained from only two sites of the interpalpebral
conjunctiva (nasal and temporal) using CIC (Hong et al., 2010). In our study, a
statistically validated approach was used to determine an acceptable level of accuracy
in the measurements of GCD at each examination.
In healthy individuals, GCD values from covered conjunctiva (upper and lower) have
been reported to be significantly higher than those from the exposed regions (nasal
and temporal) (Doughty, 2012a). However, values from the exposed conjunctiva vary
greatly from study to study. Using CIC, reports of mean GCD mean values from the
nasal bulbar conjunctiva range from 65 to 1108 cells/mm² (Kim et al., 2007;
Prabhasawat and Tseng, 1997; Rivas et al., 1991; Zhang and Yao, 2002). The reason
why these studies show such large differences in GCD values may be due to
differences in sampling techniques, such as differences in the number of images used
to report an average GCD value, the level of magnification used to image cells,
sampling area analysed, staining procedures and sample collection techniques.
Using LSCM, only one value has been reported of GCD in healthy participants, which
was from the nasal area (262 ± 116) (Efron et al., 2010a). In the present study, the
average GCD using LSCM and CIC were 475 ± 41 cells/mm² and 466 ± 51
cells/mm², respectively. The difference between studies for healthy participants could
be attributed to the validated sampling approach and the larger number of images
selected in the present study to determine GCD. As well, we adopted an image
selection criteria that required an abundant number of goblet cells to be present in
images selected for analysis (Doughty, 2012b).
Currently, assessment and quantification of GCD from images obtained by in vitro
CIC is mostly based on counts from superficial cells that easily adhere to the filter
Association between LSCM and CIC for Goblet Cell Density Assessment 109
acetate. These procedures of sample collection can result in harvesting more or less
cells depending on pressure applied to the filter and time of contact between the
ocular surface and the filter acetate.
The level of magnification and field of view used when performing microscopy
during manual cell counting can impact GCD estimates. A level of magnification of
200 x was used in this study because this magnification has been demonstrated to
introduce less variability in GC counts compared to 100 x and 400 x (Doughty,
2012b).
Given the demonstrated association between GCD measure- ments using CIC and
LSCM, researchers and clinicians may prefer to use LSCM for assessing GCD.
LSCM has the advantage of being reiterative and non-invasive. It also demonstrates
repeatable quantitative intersession measurement of cell density using cell count
software (Popper et al., 2003). Conversely, CIC is invasive (involving tissue
removal), with no evidence of repeated measure capability in the literature. Further,
repeated measurements cannot be made at the same location or region of tissue unless
a period of time is allowed for tissue regrowth.
Images obtained using LSCM can be assessed immediately, whereas a time-
consuming process of histochemical staining of CIC samples is required before cell
counts can be made. The disadvantage of LSCM is the initial cost of the
instrumentation, although CIC is also expensive when the costs of materials and
reagents is factored in as well as the time necessary for a technician to prepare, stain
and analyse the tissue samples.
In summary, we have shown that GCD assessed using CIC and LSCM are positively
correlated, meaning that either technique can be used to obtain valid results. Estimates
of GCD using LSCM can be predicted from CIC and the two methods agree. LSCM is
a relatively new approach for the assessment and quantification of goblet cells in a
non-invasive and reiterative manner, and is less time consuming than CIC.
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 111
Time Course of Changes in GCD Chapter 10:
in Symptomatic and Asymptomatic
Contact Lens Wearers
10.1 PREFACE
This Chapter addresses two main aims defined in Section 1.4 ‘Aims of the study’, of
Chapter 1 Introduction, and describes the time course of changes in GCD in
individuals symptomatic and asymptomatic of CL-induced DE over a 6-month period.
This Chapter also addresses changes in GCD of non-contact lens wearing control
participants over the same time frame. Additionally, association between LSCM and
CIC for GC assessment was also examined for longitudinal measurements.
10.2 PURPOSE
This study investigates longitudinal changes of GCD in participants who developed
symptoms of CL-induced DE compared to those who did not develop symptoms and a
non-CL wearing control group, using both LSCM and CIC.
10.3 METHODS
For the purpose of this study, the ‘General methodology and research plan’ described
in Chapter 8 was applied.
10.4 PARTICIPANTS
This prospective, longitudinal, observational, case-controlled clinical study was
conducted following approval from the QUT Human Research Ethics Committee.
Based on inclusion/exclusion criteria (Chapter 8 Sections 8.4.1 ‘Inclusion criteria’ and
8.4.2 ‘Exclusion criteria’), a group of 60 healthy individuals fitted with conventional
hydrogels worn on a daily wear basis and 23 age-balanced controls underwent
detailed assessment of DE signs and symptoms over a 6-month period.
Questionnaires DEQ-5 for controls and CLDEQ-8 for CL wearers were applied and
participants also underwent DE tests and GCD assessment using LSCM and CIC at
baseline, 1 week, 1 month, and 6 months (four time-points in total and approximately
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 112
380 case visits, including CL fitting visits). Participants excluded at baseline and
unassigned into a group at 1 week, are described in Section 8.4 ‘Study population’ of
Chapter 8. After the 1-week visit, 1, 3, and 5 participants withdrew from the control,
asymptomatic, and symptomatic groups, respectively. Then, after the 1-month visit, 1,
3, and 3 more participants withdrew from the control, asymptomatic, and
symptomatic groups, respectively. The main reasons participants withdrew from the
study was CL discomfort (6 participants), personal decision (5 participants) and lost
to follow-up (5 participants).
10.5 STATISTICAL ANALYSIS
Normality of data was examined using the Shapiro-Wilk test. To analyse the
demographic and clinical characteristics between the controls and the asymptomatic
and symptomatic CL-induced DE groups, the variables between the baseline and final
visits were compared. DE and ocular surface assessments, including questionnaires,
were compared among the controls and the asymptomatic, and symptomatic CL-
induced DE groups and for the purpose of this comparison, parametric data were
analysed using a paired sample t-test and an independent sample t-test. Nonparametric
data were analysed using an X² test, Wilcoxon test, and Mann-Whitney U test. All
data is shown as mean ± SD.
Linear mixed model (LMM) was applied to examine the changes in GCD of the nasal
bulbar conjunctiva over time. Since changes of GCD in CL wear over time was the
main parameter of interest of this study, GCD assessed by LSCM was considered the
response variable and time was added to the model to test the linear effect of CLW on
GCD. The model contained GCD as the response variable, group (i.e., the controls
and the asymptomatic and symptomatic CL-induced DE), test (i.e., LSCM and CIC),
visit (i.e., at baseline, 1 and 6 months), group * visit; and test * visit interactions as
primary fixed effects of interest and Type III sum of square was selected. Group was
included as a time-invariant predictor to determine group differences over time.
Global values of GCD were used for this analysis (the average of 5 and 11 images for
CIC and LSCM, respectively). SPSS for Windows Version 16 (SPSS Sciences,
Chicago, IL) was used for this statistical analysis and two-tailed α = 0.05 level of
significance was applied for all analyses.
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 113
10.6 RESULTS
Clinical characteristics
Table 10.1 illustrates the clinical characteristics and demographic data of participants
who did and did not develop CL-induced DE symptoms after week of CL wear and
the controls at the baseline and final visits. The baseline measurements of the
questionnaire CLDEQ-8 were considered at 1 week after the characterization of the
symptomatic and asymptomatic groups for the CL wearers. At baseline, before
characterization, the 83 participants (36 males and 47 females) in the study had a
mean age of 30.4 ± 8.4 years and were age- and gender-balanced (p = 0.89 and p =
0.22, respectively). At baseline the mean DEQ-5 score of the control group was
significantly lower than the groups with and without symptoms of CL-induced DE (p
< 0.01). At both the 1-week and 6-month visits the CLDEQ-8 score was significantly
higher (approximately two times worse) in participants with symptoms compared to
those without symptoms of CL-induced DE (p < 0.001). No significant difference
existed between the three groups in regard to the NIBUT, OSS, and PRT test (p >
0.17) at baseline. The number of participants that attended the four visits is depicted
graphically in Figure 10.1. Altogether, 81% of the enrolled participants completed the
final visit.
Figure 10.1. Distribution and number of participants examined at four visits.
23 23 22 21
35 34 32
29
25 25
20
17
0
10
20
30
40
BL 1 Week 1 Month 6 Months
Num
ber
of
par
tici
pan
ts
Controls
Asymtomatic
Symptomatic
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 114
Table 10.1. Demographic and clinical characteristics of the participants at the baseline and after 6 months of contact lens wear.
Parameter
Baseline Month 6 Follow-up p Value
Control A
Asymptomatic B**
Symptomatic C**
Control D
Asymptomatic E
Symptomatic F
AvB AvC BvC DvE DvF EvF AvD BvE CvF
n (male/female) 23 (19/4) 35 (10/25) 25 (7/18) 21 (17/4) 29 (8/21) 17 (5/12) <0.001* <0.001* 0.963* <0.001* 0.001* 0.985* 0.884* 0.612* 0.637*
Age 30.0 ± 8.0 32.1 ± 9.8 28.5 ± 6.4 30.3 ± 8.0 31.2 ± 9.3 28.1 ± 5.9 0.347† 0.545† 0.102† ― ― ― ― ― ―
CLW (h) ― ― ― ― 9.8 ± 3.9 8.3 ± 3.9 ― ― ― ― ― 0.382† ― ― ―
DEQ-5 (0-22) 2 ± 2 4 ± 2 4 ± 2 2 ± 2 ― ― 0.012+ 0.002+ 0.151+ ― ― ― 0.642# ― ―
CLDEQ-8 (0-
37) ― 12 ± 3§ 21 ± 4§ ― 11 ± 6 20 ± 9 ― ― <0.001† ― ― <0.001† ― 0.434‡ 0.122‡
NIBUT (s) 12.7 ± 6.1 11.9 ± 5.6 13.4 ± 5.7 10.0 ± 4.3 11.0 ± 4.7 9.7 ± 5.9 0. 394+ 0.987+ 0.315+ 0.522+ 0.875+ 0.122+ 0.107# 0.527# 0.248#
OSS (0-4) 0.5 ± 0.6 0.3 ± 0.5 0.4 ± 0.5 0.3 ± 0.4 0.7 ± 0.6 1.0 ± 0.8 0.301+ 0.735+ 0.518+ 0.132+ 0.013+ 0.177+ 0.582# 0.008# 0.006#
PRT (mm/20s) 22.7 ± 8.4 19.9 ± 8.0 18.0 ± 7.6 18.8 ± 7.7 16.8 ± 8.4 9.4 ± 4.9 0.171† 0.363† 0.358† 0.389† <0.001† 0.003† 0.104‡ 0.125‡ 0.001‡
GCD - LSCM
(cells/mm²) 491 ± 43 474 ± 40 466 ± 38 482 ± 33 408 ± 50 335 ± 46 0.13† 0.04† 0.44† <0.001† <0.001† <0.001† 0.45† <0.001† <0.001†
GCD – CIC
(cells/mm²) 489 ± 47 458 ± 55 459 ± 44 457 ± 41 406 ± 50 330 ± 64 0.03† 0.05† 0.93† 0.001† <0.001† 0.001† <0.001† <0.001† <0.001†
Results are expressed as mean ± SD or counts for categorical variables and two-tailed ⍺ = 0.05 level of significance was considered for all analyses (bold)
**Groups were assigned at 1-week visit; *X² Test; † Independent sample t-test; ‡ Paired sample t-test; + Mann-Whitney; #Wilcoxon Test; § CLDEQ-8 at baseline was obtained after 1-week of CL wear
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 115
CL wearing time, as shown in Table 10.1, after 6 month wear was not statistically
significantly different between symptomatic and asymptomatic groups (mean
difference 1.5 hours, p = 0.38). The symptom score (DEQ-5) in the control group was
similar at baseline compared to the 6-month visit (p = 0.64), whereas in both the
symptomatic and asymptomatic groups the symptom score (CLDEQ-8) were similar
at the baseline relative to the final visit. NIBUT was not different at the final visit
compared to baseline for the three groups (p > 0.10). OSS scores at baseline were
lower than at the final visit in the symptomatic and asymptomatic groups (p < 0.01).
The PRT scores of the symptomatic group were significantly decreased at final visit
compared to the baseline visit (p = 0.001). A significant decrease was noted in PRT
test values at the final visit of controls compared to the symptomatic group (p <
0.001). The comparison of symptomatic and asymptomatic groups also showed a
significant difference in scores of PRT at final visit (p = 0.003).
Goblet cell density
Figure 10.2 illustrates the 6-month time-course of GCD in the three groups assessed
with LSCM (A) and CIC (B).
There was more than 2-fold decrease of GCD at the 1-week visit using LSCM and
CIC in the three groups. The reduction of GCD in non-CL wearers is thought to be an
artefact caused by the removal of superficial cell layers in the nasal bulbar
conjunctiva using the CIC technique at baseline. For this reason, it could not be
concluded that GCD in the CL groups is reduced at 1 week of CLW because of the
influence of the CLs. The conjunctival epithelium is known to have a rapid healing
response by the migration of cells and mitosis. First, superficial basal cells migrate to
cover the wound. Second, the basal cells release their desmosome, creating a type of
junctional complex with neighbour cells. The normal thickness of the conjunctival
epithelium can restore within 48 to 72 hours (Kinoshita et al., 1982). However, the
time of GCD regeneration or migration from inner layers is unknown. Therefore, to
analyse the longitudinal effect of CLW on conjunctival GCD, the 1-week visit was
removed from the analysis and LMM was applied. Figure 10.3 shows the longitudinal
course of GCD over a 6-month period, excluding the 1-week visit in CL wearers and
controls assessed with LSCM and CIC.
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 116
A
B
Figure 10.2. Line graph of the longitudinal course of goblet cell density over a 6-
month period in CL wearers and controls assessed with (A) laser scanning confocal
microscopy and (B) conjunctival impression cytology. On each graph, the green line
represents the asymptomatic CL-induced DE group, the red line the symptomatic CL-
induced DE group and the blue line the control participants. Error bars indicate mean
± SD.
250
300
350
400
450
500
550
-1 0 1 2 3 4 5 6
GC
D c
ells
/mm
²
Months
100
175
250
325
400
475
550
-1 0 1 2 3 4 5 6
GC
D c
ells
/mm
²
Months
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 117
A
B
Figure 10.3. Line graph of the longitudinal course of goblet cell density over a 6-
month period, excluding the 1-week visit, in CL wearers and controls assessed with
(A) laser scanning confocal microscopy and (B) conjunctival impression cytology. On
each graph the green line represents the asymptomatic CL-induced DE group, the red
line symptomatic CL-induced DE group, and the blue line the control participants.
Error bars indicate mean ± SD.
250
300
350
400
450
500
550
-1 0 1 2 3 4 5 6
GC
D c
ells
/mm
²
Months
250
300
350
400
450
500
550
-1 0 1 2 3 4 5 6
GC
D c
ells
/mm
²
Months
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 118
To understand the effect of symptom grouping and time on the outcome variation, the
LLM was applied (Table 10.2). The Type III test of fixed effects shows overall
significance for the predictor variables. There was a significant effect of group and
visit; however, the effect of test (LSCM vs CIC) was not significant. The interaction
between groups and visits was significant and no significant interaction existed
between test and visit.
Table 10.2. The effect of time (visit), group and test on the dependent variable GCD
assessed with LSCM using the type III of fixed effects from LMM analysis.
LMM
F p
Intercept 5991.2 <0.001
Visit 78.4 <0.001
Test 0.8 0.36
Group 3.7 <0.001
Group * Visit 12.1 <0.001
Test * Visit 1.1 0.30
A second subset of fixed effect parameters for GCD assessed with LSCM as the
continuous response variable was included in the LMM. Parameters estimates and
corresponding 95% confidence intervals (CI), standard errors (SE), and p values are
given in Table 10.3. Regardless of group, GCD was not influenced by test at the 1-
month visit and did not show a significant change (β = -3.08, p = 0.80), whereas at the
6 month visit, was significant (β = -29.73, p <0.001). In the model, there was no effect
of test (LSCM vs CIC) on GCD (p = 0.36), however, there was a significant effect of
group (p < 0.000). At the 1-month visit, the interaction between group and visit was
significant for CL wearers compared to controls (p < 0.05). The LMM also showed a
differential effect of time on the GCD with a decrease of 127.86 cells/mm² in
individuals who developed symptoms for DE induced by CLW and 84.44 cells/mm²
in the asymptomatic group compare with controls (reference group). There was not
effect of test, and the interaction between test and visit was no significant (p = 0.30).
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 119
Table 10.3. Maximum likelihood of the fixed effect parameters for LMM with GCD
assessed with LSCM as the continuous response variable.
Parameter Estimate (95% CI) SE p
Intercept 472.65 (446.99 to 498.31) 12.39 <0.000
Visit Visit 1M -3.08 (-26.65 to -20.50) 11.98 0.80
Visit 6M -29.73 (-53.03 to 6.44) 11.85 <0.000
Visit BL 0* Test
LSCM -6.27 (-23.64 to 11.09) 8.84 0.36
CIC 0* Group
Symptomatic -127.86 (-163.36 to -92.37) 11.68 <0.000
Asymptomatic -84.44 (-115.45 to -53.44) 10.68 <0.000
Controls 0* Group*Visit
Symptomatic * Visit 1M -41.25 (-84.86 to -13.64) 18.11 0.007
Asymptomatic * Visit 1M -30.81 (-61.49 to 0.11) 15.60 0.049
Controls * Visit 1M 0* Symptomatic * Visit 6M -100.68 (-135.41 to -65.94) 17.66 <0.000
Asymptomatic * Visit 6M -60.84 (-91.08 to -30.60) 15.38 <0.000
Controls * Visit 6M 0* Test*Visit
LSCM * Visit -3.00 (-8.20 to 2.21) 2.65 0.30
CIC * Visit 0* CI, confidence interval
10.7 DISCUSSION
In vivo assessment of conjunctival GCD using LSCM has emerged as a modern non-
invasive technique that improves the investigation, not only in CL-induced DE
symptoms, but also in any CL-induced changes in the anterior eye (Efron, 2007). As
reviewed previously, GCD assessment using LSCM has been discussed in ocular
surface changes related to age (Wei et al., 2011; Zhu et al., 2010), Sjögren’s
syndrome DE (Hong et al., 2010; Villani et al., 2011a), pterygium (Labbé et al.,
2010), chemical burns (Le et al., 2010), glaucoma treatment with preserved and
unpreserved levobunolol (Ciancaglini et al., 2008) and tafluprost therapy
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 120
(Mastropasqua et al., 2013). To date, only one cross-sectional report of GCD assessed
by LSCM in CLW showed no significant changes between CL wearers and non-CL
wearers; however, this study was limited by small sample size (Efron et al., 2009).
Over a 6-months period, GCD was reduced by approximately 13% in the
asymptomatic and 29% in the symptomatic group. There was not significant
difference between LSCM and CIC over time (p = 0.30). The control group remained
relatively constant over time with a coefficient of variation of 4% and 7% using
LSCM and CIC, respectively.
This study reports GC changes in a relatively large sample of healthy individuals (n =
60) fitted with daily disposable conventional hydrogels, and healthy non-CL wearers
(n = 23) assessed over a 6-month period. Previous longitudinal assessments of GCD
in CLW have been undertaken using ex vivo analysis with CIC (Connor et al., 1997;
Connor et al., 1994; Lievens et al., 2003; Simon et al., 2002). The present study
examines, for the first time, longitudinal changes of GCD in CL wearers using both
LSCM and the well-stablished CIC technique.
It was important to ensure all individuals at baseline were free from signs and
symptoms of DE so that the chnages observed could be attributed to the CLW. At
baseline all participants had normal scores in the DEQ-5 questionnaire and were
within normal limits for at least one of the three ocular surface assessments (NIBUT,
PRT, and OSS) in accordance with the DEWS guidelines (Report of the International
Dry Eye WorkShop (DEWS), 2007). At the 1-week visit, 42% the wearing CLs were
categorised as having CL-induced DE group with scores of 17 or more points in the
CLDEQ-8 questionnaire. The remaining 58% were asymptomatic of CL-included DE
with scores of 16 or less. Age was similar between the CL and non-CL wearing
groups to ensure any age effect.
Differences were noted between the symptomatic, asymptomatic and control groups
although all the participants had passed the DEQ-5 at the baseline visit with scores of
6 points or less. This difference in the scores could be attributed to the fact that this
questionnaire was developed for non-CL wearers, which could potentially develop
DE in CLW based on frequency and intensity of the symptoms. This pattern was
clearly reflected in the three groups before developing symptoms of DE-induced by
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 121
CLW. The mean and SD of the controls and the asymptomatic and symptomatic
groups showed scores of 2 ± 2, 4 ± 2, and 4 ± 2, respectively, and the comparison
between controls and CL groups was statistically significant (p < 0.05).
A possible reason for the significant difference in the scores of DEQ-5 within the CL
groups and controls could be that individuals with refractive errors are more likely to
self-report DE symptoms than emmetropes (Nichols et al., 2005). In this study, all the
participants in the control group were emmetropes except for 1 participant with
presbyopia who was not interested in wearing CL and who showed scores of 3 points
in the DEQ-5. All the CL wearers have refractive errors between -6.75 to +1.75 D and
astigmatism from -0.25 to -1.25 D. Comparison of the NIBUT, OSS, and PRT at the
baseline visit showed no significant difference in the signs of DE and OSS in any of
the groups, showing normal scores and absence of DE.
Comparison of CLDEQ-8 scores between symptomatic and asymptomatic groups at
the 1-week and 6-month visit showed significant differences to the self-reported DE
symptoms, indicating desirable differences for the characterization between these two
groups. NIBUT showed no statistically significant changes in any group when
comparing the baseline to the final visit (p > 0.05). The mechanical effect of lens
removal before the assessment may be a confounding factor of tear film stability. OSS
was increased significantly in symptomatic and asymptomatic participants at the
baseline and final visits, which was an expected effect of the CL. Although the
symptomatic group showed significantly lower measurements of PRT at the final
visit, the asymptomatic group did not. PRT was found to have high repeatability on
the same group of participants assessed in previous studies (Cho, 1993; Little and
Bruce, 1994). However, there is no evidence that PRT measurements are sensitive
enough to differentiate from mild to moderate tear volume values. This would be
necessary to stablish in studies where the severity of symptoms are measured using
this test.
Based on the results of the majority of the studies previously mentioned in regard to
the impact of CLW on conjunctival GCs, it was hypothesised a reduction of GCD as a
result of CLW. To examine this hypothesis, a LMM was developed. This model
afforded the analysis of longitudinal data with repeated measures for samples with
unequal variances because of missing data and dropouts.
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 122
The LMM revealed that regardless of group, the test had no significant effect on
GCD. A test * time interaction term was not significant (p = 0.30), indicating that the
assessment of GCD using LSCM and CIC did not differ over time regardless of the
changes in GCD. A group * visit interaction showed a decrease of GCD from the
baseline to the 1-month visit in the symptomatic and asymptomatic groups compared
to the controls of 41.25 and 30.81 cells/mm², respectively. A further reduction of
100.68 and 60.84 cells/mm² was observed in the symptomatic and asymptomatic
groups, respectively, from the baseline to the final visit. This further reduction
suggests that there was an accelerated decrease of GCD from baseline to 1 month
compared to the GCD reduction from the 1-month to final visit. This reduction of
GCD was approximately 59.43 and 30.03 cells/mm² in the symptomatic and
asymptomatic group, respectively. An additional model was developed in order to
determine longitudinal change only for the control group. The Type III fixed effects
showed no significant changes from the baseline to final visit (p = 0.14), suggesting
that the GCD decreased in the symptomatic and asymptomatic groups due to the
effect of CLW.
The results of this study disagree with the results from Connor (1994; 1997), Lievens
(2003), and Corrales (2009), who reported a statistically significant increase or no
changes to GCD assessed with CIC over a 6-month and 1-year period in participants
fitted with CLs. This disagreement may be attributed to the variances in the following
factors relating to CIC technique: sample inconsistency across the filter; number of
images sampled; criteria for identifying GCs; quality of acetate filter used for cell
attachment; units used to report GCD; and conjunctival region assessed. Also, lens-
related factors such as material (e.g., conventional hydrogels versus silicone
hydrogels) and replacement frequency (e.g., daily/monthly replacement) make it
difficult to compare results of various studies. Few of the previous studies (Connor et
al., 1997; Connor et al., 1994; Lievens et al., 2003) incorporated a non-CL wearing
control group.
Corrales (2009) tried to avoid the variable of sampling inconsistency by using
biomarker for mucus detection (MUC5AC) isolated from GCs assessed with CIC and
analysed with PCR (Corrales et al., 2009). This study, however, associated the
changes in GCD with the expression of MUC5AC detected after 1 year of 66% water
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 123
content CL wear compared to 38% water content CL wear (p = 0.02), and the values
of mucin were given in a logarithmic unit (log 0.8). Connor (1994) reported a 2-fold
increase in GCD of 18 participants fitted with 38% water content CL wear on a daily
basis, without CL replacement, initially for 6 months. CIC was performed on the
lower bulbar, and only one random image per sample was used to report a percentage
of GCs in the total cell population; GCD was given in a percentage unit out of the
total number of cells in the field of view including ephitelial cells (8.08%). Connor
and co-workers concluded that the 2-fold increase was due to lipid deposits on the
lenses after 6 months of CLW (Connor et al., 1994).
In 1997, the same authors duplicated the same study using a bigger sample size (n =
28) and rather than CL replacement every 6 months, replaced lenses every 2 weeks. In
this study, Connor and co-workers found no significant changes from the baseline to
final visit and concluded that frequent replacement was less irritating, thus influencing
the GC response (Connor et al., 1997).
In 2003, Lievens in association with Connor repeated the study using two different
silicone hydrogels in extended wear modality (6 consecutive days) with weekly and
monthly lens replacement. At this time, a significant increase was found with no
difference between the two groups and a conclusion was drawn by these investigators
that silicone hydrogels were promising extended wear lenses due to the their lack of
impact on GCD.
There are further methodologic problems with the three above-mentioned studies.
These authors used only one image per sample to determine GCD. The GCD
calculation was expressed as the percentage of GCs in the total number of cells
counted per field of view (including non-goblet cells). In order to have GCs attach to
an acetate filter, the sample must be multilayered (Albietz, 2001; Colorado et al.,
2016; Doughty, 2012b). This is because GCs are about three times bigger than
surrounding cells, and they need support from the epithelial non-goblet cells in order
to attach in the filter as mentioned before by many authors and experienced in this
present study. If a multilayered sample is stained with PAS for GC detection and
counterstained for non-goblet cells, it is not possible to undertake a total cell count
due to the two to three non-goblet overlapping cells obscuring the view. In these
papers, there is a lack of evidence of the dimensions of the field of view and
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 124
magnification used for observation and quantification of the cells. Without this
information, the results are difficult to interpret. There is no clear explanation why
GC increased and did not change after CLW in the above-mentioned studies.
In contrast, the present study agrees with the longitudinal findings of Simon et al.
(2002), who reported a decrease of GCD after fitting participants with soft and rigid
CLs for 6 months. These authors used CIC and correlated the severity of symptoms
with cytological alterations using a grading system for squamous metaplasia (Simon
et al., 2002). In this study, results of GCD are given using a specific value in units of
cells/mm² because categorising a continuous predictor could result in loss of statistical
power (Aiken et al., 1991).
In this longitudinal study, the relationship between CLW-induced DE symptoms and
GCD was examined using the conventional CIC method and the promising LSCM for
GC assessment. The findings of this study demonstrate a longitudinal equivalence
between LSCM and CIC. The strengths of this study are the recruitment of sufficient
participants as guided by power analysis; use of both a non-invasive, reiterative
technique (LSCM) and a gold standard invasive technique (CIC) to assess GCD;
careful phenotyping of symptomatic and asymptomatic lens wearers; incorporation of
a non-CLW control group; appropriate masking of observers; use of validated image
capture (LSCM) and sample collection (CIC) protocols; masked image selection and
analysis methodology, and the use of robust statistical modelling.
A general limitation of all studies using CIC is the lack of evidence relating to the
validity of undertaking repeated measures. We have demonstrated that it may take up
to 4 weeks for GCD to recover post-CIC using Biopore Millicell inserts. The
reduction of GCD in non-CL wearers at 1 week is attributed to an artefact caused by
the removal of superficial cell layers of the epithelium. The conjunctival epithelium is
known to have a rapid healing response, effected by cell migration and mitosis,
whereby normal thickness is restore within 48 to 72 hours (Kinoshita et al., 1982).
However, the time-course of GC differentiation, regeneration or migration from inner
layers is unknown. It is also unknown whether CL prevents healing as fast post‐CIC
than the normal eye. The recovery time for the repopulation of GCs after having been
removed from the conjunctival surface with different types of acetate filters is also
unknown.
Time Course of Changes in GCD in Symptomatic and Asymptomatic Contact Lens Wearers 125
In summary, the results of this study provide evidence that LSCM has the capability
to detect changes of GCD in patients with CL-induced DE. Furthermore, GCD is
reduced in CLW, and further reductions are found in those who develop symptoms.
Summary, Conclusions and Recommendations 127
Summary, Conclusions and Chapter 11:
Recommendations
11.1 SUMMARY OF THE PROJECT OUTCOMES
The thesis explores the time course of changes in GCD in response to symptoms of
CL-induced DE, comparing in vivo non-invasive and ex vivo mildly-invasive
techniques of assessment. Identifying early changes of GCD in conjunction with
validated DE tests will inform those developing strategies for successful and
comfortable CLW. The refinement and cross-validation of repeatable methods for the
GC assessment represents a significant advance in the assessment of CL-induced DE.
The thesis examines the association between in vivo LSCM and ex vivo CIC as an
assessment technique for conjunctival GCD, and determines longitudinal changes of
GCD in symptomatic and asymptomatic CL-induced dry eye—a multifactorial
complication of CLW that currently affects approximately 50% of CL wearers for
which there is no effective treatment.
The two methods used to assess GCD in this research project were carefully
characterized, and validation studies were undertaken using immunohistochemical
and immunocytochemical techniques. This thesis also explores the influence of test
order and repeatability when determining GCD. An image sampling analysis
technique was devised to avoid measurement bias when assessing GC images
obtained using LSCM and CIC.
In the human body, each cell type has unique biomarkers that facilitate the
characterization of a cell genotype, differentiation, and isolation that can be analyzed
using sophisticated and expensive methods such as immunofluorescence staining,
PCR, and FC from isolate cell assessed by CIC. However, there are cell types that can
be easily recognized by morphological appearance according to the tissue location
and morphology of surrounding cells. For example, conjunctival GCs are found
scattered among the epithelial lining of the conjunctival epithelium, having a height of
three to four times that of their width, with a distinct balloon-like appearance.
Summary, Conclusions and Recommendations 128
The structures presumed to be conjunctival GCs observed using LSCM have not, to
date, been verified as such. The objective of the tissue experiment presented in
Chapter 3 was to verify that features previously identified as GCs using LSCM are
truly the same GCs identified using antibody biomarkers for cell recognition. This
experiment provided IHC evidence that features observed using LSCM that are
presumed to be GCs are most likely this cell type. Further investigations along these
lines should include examinations of conjunctival tissue from enucleated eyes or other
conjunctival surgeries.
Immunocytochemical identification of GCs using PAS has been the approach
favoured for previous authors to detect mucoproteins in cytological samples, probably
because PAS stain highlights the presence of mucus cells. Chapter 4 described an
alternative approach by investigating GCD by immunocytochemical and
morphological identification using Giemsa stain and comparing the outcome with that
obtained using PAS. This experiment demonstrated that Giemsa stain has the
advantage over PAS of facilitating visualization of cell borders and cell nuclei.
Giemsa staining is also more time- and cost-effective staining technique than PAS,
making cell counting more reliable in CIC specimens.
LSCM was explored as an alternative method for quantifying GCD as this technique
is less invasive than CIC. The reproducibility of this technique was explored in
Chapter 5 using Bland-Altman repeatability (Bland and Altman, 1986), and the intra-
class correlation coefficients for one observer measuring GCD on two different
occasions. It was demonstrated that conjunctival GCD assessed by LSCM could be
measured in a repeatable manner. The findings of this study showed that LSCM
achieved a reasonably acceptable repeatability level for determination of GCD.
However, it is acknowledged that small intra-observer differences in GCD assessed
using LSCM would be difficult to detect as demonstrated in this chapter.
The use of CIC after LSCM assessment on the same area of tissue – in this case
bulbar conjunctiva – has been used by previous authors to compare the techniques
with regard to GCD related to epithelial cell changes (squamous metaplasia).
However, Mastropasqua et al. (2013), who separated the LSCM and CIC
measurements by 24 hours, indicate that the results of GCD assessed by CIC could
potentially be affected by previous assessments using LSCM. The purpose of the
Summary, Conclusions and Recommendations 129
experiment presented in Chapter 6 was to investigate whether GCD measured with
CIC is compromised by prior LSCM examination. This study demonstrated that there
are no significant effects on GCD measured using CIC after LSCM examination.
Therefore, in the main longitudinal study, the GC assessment was performed using
CIC a few minutes after the LSCM examination.
To avoid measurement bias when establishing GCD using LSCM and CIC techniques,
a method to quantify a number of random GC images was addressed in Chapter 7.
The number of images necessary for achieving an acceptable level of accuracy for
determining GCD using LSCM and CIC was eleven and 5, respectively.
In order to answer the two main research questions of this study described in Chapters
9 and 10, the results of the studies previously mentioned were appraised so as to
inform the general methodology and research plan developed for Chapter 8.
To explore the association between in vivo LSCM and ex vivo CIC as an assessment
technique for conjunctival GCD, a regression approach for non-uniform differences
was carried out using Bland-Altman technique with linear regression and 95% LoA
(Bland et al., 1999) (Chapter 9). The image sampling analysis described in Chapter 7
was implemented for this analysis (the average of 5 and eleven images for CIC and
LSCM, respectively). This study demonstrated that GCD assessed using CIC and
LSCM are highly correlated, meaning that either technique can be used to obtain valid
results. The measurements of GCD assessed by LSCM can be predicted from those by
CIC and the two methods agree. Demonstration of an association between these two
techniques will serve to validate LSCM as a viable alternative procedure to assessing
GCD in human populations.
Chapter 10 reports an experiment that set out to determine the time course of changes
in GCD in symptomatic and asymptomatic CL wearers (i.e. those with and without
CL-induced DE) using LSCM and CIC. A control group was also assessed. GCD is
reduced in CLW, and further reductions are found in those who develop symptoms.
Furthermore, the results of this study provide evidence that LSCM has the capability
to detect changes of GCD in patients with CL-induced DE.
Summary, Conclusions and Recommendations 130
11.2 CONTRIBUTION TO NEW KNOWLEDGE
This study attempted to resolve discrepancies in the literature concerning the
time course of changes of GCD in CLW using validated, methodological
procedures. The most significant contributions of this study include:
Providing evidence that LSCM has the capability and repeatability to measure
GCD as an alternative to CIC;
Confirming the impact of CL wear on GCD in symptomatic and asymptomatic
patients; and
Demonstrating that LSCM not only allows the visualization of GCs and their
distribution in the conjunctiva, but is also capable of observing GCD changes
non-invasively over time in CL-induced DE and in healthy participants.
11.3 RECOMMENDATIONS FOR FUTURE RESEARCH
The research study presented in this thesis provides the first evidence of longitudinal
changes in GCD related to CL-induced DE measured using a non-invasive technique,
LSCM. However, more research is required to establish this technique as a validated
clinical tool.
In this study, conjunctival GCD was shown to be reduced to a greater extent in
individuals with symptoms of CL-induced DE. Therefore, conjunctival GCD may be
an important marker of CL discomfort. Both ex vivo and in vivo techniques
demonstrated this phenomenon.
Further validation of CIC as an assessment technique of GCD could be conducted if it
is to be used in the future. Marked differences in GCD estimates from CIC specimens
reported previously could be due in part to variability in the results of earlier studies
that could be related to a number of factors, including differences in sampling
location, quality of acetate filters, the grading scale used to report squamous
metaplasia, and magnification and field of view used to examine the specimen.
In this study, the repeatability of LSCM as an assessment technique for GCD was
demonstrated, and similar findings by other authors with regard to GCD in healthy
participants have been reported (Zhu et al., 2010). Therefore, LSCM could replace
CIC as a less invasive and more reliable method for GC assessment. To understand
Summary, Conclusions and Recommendations 131
the effect of CLW on GCD, further investigation needs to be conducted using this
technique to address changes of GCD with different contact lens materials, lens
designs, lens replacement frequencies, wearing modalities (extended vs daily wear)
and lens care systems. In addition, recent evidence highlighting the importance of the
lid wiper and its role in contact lens comfort indicates that new studies should be done
measuring GCD in this tissue because, as demonstrated in this body of work, GCD
may play a key role in CL discomfort and may be considered as an appropriate
marker for future studies.
These studies are the first attempt to employ LSCM as an assessment technique for
GCD in a longitudinal manner. Further efforts to use LSCM as a first-choice
assessment method of GCD should be made, perhaps to extend the advantages of this
technique as a clinical and research tool with the purpose to find possible
management strategies for alleviating CL discomfort, the main reason for CL
discontinuation.
A possible implication of GC loss in CL wearers is the compromise in the production
of the mucous-secreting component of the tear film which may result in CL
intolerance and the development of complications of longer term CL wear.
Preventative measures of GCD using LSCM in combination with symptomatology
gathered from validated questionnaires are recommended to be adopted in clinical
practice before and after fitting patients with CLs.
Practitioners could prevent patients from suffering a loss of goblet cells by
determining the best way to increase the comfort of wearing contact lenses and to
reduce symptomatology. Today, a variety of new contact lenses and care products
have been made in order to reduce discontinuation of contact lens wear. Silicone
hydrogel lenses in single-use daily bases accompanied by non-preservative lubricant
eye drops should be recommended as supplemental measures to reduce contact lens
discomfort. Alternatively, in terms of daily schedule, replacement of the contact
lenses every 6 hours with no more than 12 hours of continuous wear could improve
symptomatology and discomfort. However, there is evidence that alternative therapies
such as Omega 3 fatty acids (specifically docosahexaenoic acid, eicosapentaenoic
acid, and alpha-linolenic acid) may also benefit eye health. Therefore, this measure
could also be adopted for some patients.
Bibliography 133
Bibliography
Adams, G., Dilly, P.,Kirkness, C. (1988). Monitoring ocular disease by impression
cytology. Eye, 2(5), 506-516.
Adar, S., Kanpolat, A., Sürücü, S.,Ömür Ucakhan, Ö. (1997). Conjunctival
impression cytology in patients wearing contact lenses. Cornea, 16(3), 289-
294.
Aiken, L. S., West, S. G.,Reno, R. R. (1991). Reliability and Statistical Power. In S.
McElroy (Ed.), Multiple regression: Testing and interpreting interactions (pp.
156-170). Thousand Oaks, CA, US: Sage Publications.
Aksünger, A., Ünlü, K., Karakaş, N., Nergiz, Y.,Çelik, Y. (1997). Impression
cytology, tear film break up, and Schirmer test in patients with inactive
trachoma. Japanese journal of ophthalmology, 41(5), 305-307.
Albietz, J., Sanfilippo, P., Troutbeck, R.,Lenton, L. M. (2003). Management of
filamentary keratitis associated with aqueous-deficient dry eye. Optometry &
Vision Science, 80(6), 420-430.
Albietz, J. M. (1999). The conjunctiva in dry eye : histological changes,
inflammation, topical treatments, contact lenses and refractive surgery
(Doctoral dissertation). Queensland University of Technology, Brisbane,
QLD.
Albietz, J. M. (2001). Conjunctival histologic findings of dry eye and non-dry eye
contact lens wearing subjects. Eye & Contact Lens, 27(1), 35-40.
Anshu, Munshi, M., Sathe, V.,Ganar, A. (2001). Conjunctival impression cytology in
contact lens wearers. Cytopathology, 12(5), 314-320.
Aragona, P., Ferreri, G., Micali, A.,Puzzolo, D. (1998). Morphological changes of the
conjunctival epithelium in contact lens wearers evaluated by impression
cytology. Eye, 12, 461-466.
Argüeso, P., Balaram, M., Spurr-Michaud, S., et al. (2002). Decreased levels of the
goblet cell mucin MUC5AC in tears of patients with Sjögren syndrome.
Investigative ophthalmology & visual science, 43(4), 1004-1011.
Argüeso, P.,Gipson, I. K. (2001). Epithelial mucins of the ocular surface: structure,
biosynthesis and function. Experimental Eye Research, 73(3), 281-289.
Arita, R., Itoh, K., Inoue, K., et al. (2009). Contact lens wear is associated with
decrease of meibomian glands. Ophthalmology, 116(3), 379-384.
Bibliography 134
Bai, T., Huang, J.,Wang, W. (2010). Short-term comparative study of the effects of
preserved and unpreserved topical levofloxacin on the human ocular surface.
Cutaneous and ocular toxicology, 29(4), 247-253.
Barr, J. (2005). 2004 Annual report. Contact Lens Spectrum, 20(1), 26-32.
Baudouin, C., Aragona, P., Messmer, E. M., et al. (2013). Role of hyperosmolarity in
the pathogenesis and management of dry eye disease: proceedings of the
OCEAN group meeting. The ocular surface, 11(4), 246-258.
Begley, C. G., Chalmers, R. L., Abetz, L., et al. (2003). The relationship between
habitual patient-reported symptoms and clinical signs among patients with dry
eye of varying severity. Investigative ophthalmology & visual science, 44(11),
4753-4761.
Bjerrum, K. B. (1996). Test and symptoms in keratoconjunctivitis sicca and their
correlation. Acta Ophthalmologica Scandinavica, 74(5), 436-441.
Blades, K. J.,Patel, S. (1996). The dynamics of tear flow within a phenol red
impregnated thread. Ophthalmic and Physiological Optics, 16(5), 409-415.
Bland, J. M.,Altman, D. (1986). Statistical methods for assessing agreement between
two methods of clinical measurement. The lancet, 327(8476), 307-310.
Bland, J. M.,Altman, D. G. (1999). Measuring agreement in method comparison
studies. Statistical methods in medical research, 8(2), 135-160.
Bourne, W. (2003). Biology of the corneal endothelium in health and disease. Eye,
17(8), 912-918.
Bourne, W. M. (1983). Morphologic and functional evaluation of the endothelium of
transplanted human corneas. Transactions of the American Ophthalmological
Society, 81, 403-450.
Brennan, N. A., Coles, M. C., Connor, H. R., et al. (2008). Short-term corneal
endothelial response to wear of silicone-hydrogel contact lenses in East Asian
eyes. Eye & contact lens, 34(6), 317-321.
Bron, A., Tiffany, J., Gouveia, S., Yokoi, N.,Voon, L. (2004). Functional aspects of
the tear film lipid layer. Experimental eye research, 78(3), 347-360.
Bron, A. J. (2001). Diagnosis of dry eye. Survey of Ophthalmology, 45(2), S221-
S226.
Brooks, A., Grant, G., Westmore, R.,Robertson, I. F. (1986). Deep corneal stromal
opacities with contact lenses. Australian and New Zealand journal of
ophthalmology, 14(3), 243-249.
Brücke, E. W. (1847). Anatomische Beschreibung des menschlichen Augapfels.
Berlin: Reimer.
Bibliography 135
Çakmak, S. S., Ünlü, M. K., Karaca, C., Nergiz, Y.,Ipek, S. (2003). Effects of soft
contact lenses on conjunctival surface. Eye & Contact Lens, 29(4), 230-233.
Carnt, N. A., Evans, V. E., Naduvilath, T. J., et al. (2009). Contact lens–related
adverse events and the silicone hydrogel lenses and daily wear care system
used. Archives of ophthalmology, 127(12), 1616-1623.
Cavanagh, H. D., Ladage, P. M., Li, S. L., et al. (2002). Effects of daily and overnight
wear of a novel hyper oxygen-transmissible soft contact lens on bacterial
binding and corneal epithelium: a 13-month clinical trial. Ophthalmology,
109(11), 1957-1969.
Chalmers, R. L., Begley, C. G.,Caffery, B. (2010). Validation of the 5-Item Dry Eye
Questionnaire (DEQ-5): Discrimination across self-assessed severity and
aqueous tear deficient dry eye diagnoses. Contact Lens and Anterior Eye,
33(2), 55-60.
Chalmers, R. L., Begley, C. G., Moody, K.,Hickson-Curran, S. B. (2012a). Contact
Lens Dry Eye Questionnaire-8 (CLDEQ-8) and Opinion of Contact Lens
Performance. Optometry & Vision Science, 80(10), 1435-1442.
Chalmers, R. L., Hunt, C., Hickson-Curran, S.,Young, G. (2009). Struggle with
hydrogel CL wear increases with age in young adults. Contact Lens and
Anterior Eye, 32(3), 113-119.
Chalmers, R. L., Keay, L., McNally, J.,Kern, J. (2012b). Multicenter case-control
study of the role of lens materials and care products on the development of
corneal infiltrates. Optometry & Vision Science, 89(3), 316-325.
Chan, C. M., Liu, Y. P.,Tan, D. T. (2002). Ocular surface changes in pterygium.
Cornea, 21(1), 38-42.
Chang, S. W., Hu, F. R.,Lin, L. (2000). Effects of contact lenses on corneal
endothelium-a morphological and functional study. Ophthalmologica, 215(3),
197-203.
Chen, H. B., Yamabayashi, S., Ou, B., et al. (1997). Structure and composition of rat
precorneal tear film. A study by an in vivo cryofixation. Investigative
ophthalmology & visual science, 38(2), 381-387.
Chew, C., Jansweijer, C., Tiffany, J., Dikstein, S.,Bron, A. (1993). An instrument for
quantifying meibomian lipid on the lid margin: the Meibometer. Current Eye
Research, 12(3), 247-254.
Cho, P. (1993). The cotton thread test: a brief review and a clinical study of its
reliability on Hong Kong-Chinese. Optometry & Vision Science, 70(10), 804-
808.
Bibliography 136
Cho, P.,Douthwaite, W. (1995). The relation between invasive and noninvasive tear
break-up time. Optometry & Vision Science, 72(1), 17-22.
Ciaccio, V. (1873). Osservazioni intorno alla struttura della congiuntiva umana (Vol.
4). Bologna, Italy: Academy of Sciences of the Institute of Bologna.
Ciancaglini, M., Carpineto, P., Agnifili, L., et al. (2008). An in vivo confocal
microscopy and impression cytology analysis of preserved and unpreserved
levobunolol-induced conjunctival changes. European journal of
ophthalmology, 18(3), 400-407.
Colorado, L. H., Alzahrani, Y., Pritchard, N.,Efron, N. (2016). Assessment of
conjunctival goblet cell density using laser scanning confocal microscopy
versus impression cytology. Contact Lens and Anterior Eye, 39(3), 221-226.
Connor, C., Campbell, J.,Steel, S. (1997). The effects of disposable daily wear contact
lenses on goblet cell count. Contact Lens Association of Ophthalmologists,
23(1), 37-39.
Connor, C., Campbell, J., Tirey, W., Steel, S.,Burke, J. (1991). Modification of
impression cytology for in-office use. Journal of the American Optometric
Association, 62(12), 898.
Connor, C. G., Campbell, J. B., Steel, S. A.,Burke, J. H. (1994). The effects of daily
wear contact lenses on goblet cell density. Journal of the American
Optometric Association, 65(11), 792-794.
Corrales, R. M., Galarreta, D., Herreras, J. M., et al. (2009). Conjunctival mucin
mRNA expression in contact lens wear. Optometry & Vision Science, 86(9),
1051-1058.
Cruzat, A., Witkin, D., Baniasadi, N., et al. (2011). Inflammation and the nervous
system: the connection in the cornea in patients with infectious keratitis.
Investigative ophthalmology & visual science, 52(8), 5136-5143.
Curran, R. E., Kenvon, K. R.,Green, W. R. (1974). Pre-Descemet's membrane corneal
dystrophy. American journal of ophthalmology, 77(5), 711-716.
Dada, V., Jain, A.,Mehta, M. (1989). Specular microscopy of unilateral hard contact
lens wearers. Indian Journal of Ophthalmology, 37(1), 17-19.
Dartt, D. A., Kessler, T. L., Chung, E. H.,Zieske, J. D. (1996). Vasoactive intestinal
peptide-stimulated glycoconjugate secretion from conjunctival goblet cells.
Experimental eye research, 63(1), 27-33.
Dartt, D. A., Mccarthy, D. M., Mercer, H. J., et al. (1995). Localization of nerves
adjacent to goblet cells in rat conjunctiva. Current Eye Research, 14(11), 993-
1000.
Bibliography 137
Dawson, D. G.,Edelhauser, H. F. (2010). Corneal edema. In Ocular Disease:
Mechanisms and Management (pp. 64-73): eElsevier.
De Duve, C.,Wattiaux, R. (1966). Functions of lysosomes. Annual review of
physiology, 28(1), 435-492.
Diebold, Y., Ríos, J. D., Hodges, R. R., Rawe, I.,Dartt, D. A. (2001). Presence of
nerves and their receptors in mouse and human conjunctival goblet cells.
Investigative Ophthalmology & Visual Science, 42(10), 2270-2282.
Dilly, P. (1985). Contribution of the epithelium to the stability of the tear film.
Transactions of the Ophthalmological Societies of the United Kingdom, 104,
381-389.
Doane, M.,Gleason, W. (1994). Tear layer mechanics. In W. BA (Ed.), Clinical
Contact Lens Practice (pp. 1-17). Philadelphia: Lippincott.
Dogru, M., Karakaya, H., Özçetin, H., et al. (2003). Tear function and ocular surface
changes in keratoconus. Ophthalmology, 110(6), 1110-1118.
Dogru, M., Katakami, C.,Inoue, M. (2001). Tear function and ocular surface changes
in noninsulin-dependent diabetes mellitus. Ophthalmology, 108(3), 586-592.
Dogru, M., Katakami, C., Miyashita, M., et al. (2000). Ocular surface changes after
excimer laser phototherapeutic keratectomy1. Ophthalmology, 107(6), 1144-
1152.
Dogru, M., Katakami, C., Nakagawa, N., Tetsumoto, K.,Yamamoto, M. (1998).
Impression cytology in atopic dermatitis. Ophthalmology, 105(8), 1478-1484.
Dogru, M., Matsumoto, Y., Okada, N., et al. (2008). Alterations of the ocular surface
epithelial MUC16 and goblet cell MUC5AC in patients with atopic
keratoconjunctivitis. Allergy, 63(10), 1324-1334.
Dogru, M., Özmen, A., Ertürk, H., Sanli, Ö.,Karatas, A. (2002). Changes in tear
function and the ocular surface after topical olopatadine treatment for allergic
conjunctivitis: an open-label study. Clinical Therapeutics, 24(8), 1309-1321.
Dorland, W. A. (2012). Dorland's illustrated medical dictionary. Philadelphia, PA:
Elsevier.
Doughty, M. J. (2011a). Contact lens wear and the goblet cells of the human
conjunctiva-A review. Cont Lens Anterior Eye, 34(4), 157-163.
doi:10.1016/j.clae.2011.04.004
Doughty, M. J. (2011b). Objective assessment of contact lens wear-associated
conjunctival squamous metaplasia by linear measures of cell size, shape and
nucleus-to-cytoplasm ratios. Current Eye Research, 36(7), 1-8.
Bibliography 138
Doughty, M. J. (2012a). Goblet cells of the normal human bulbar conjunctiva and
their assessment by impression cytology sampling. The Ocular Surface, 10(3),
149-169.
Doughty, M. J. (2012b). Sampling area selection for the assessment of goblet cell
density from conjunctival impression cytology specimens. Eye and Contact
Lens, 38(2), 122-129.
Doughty, M. J., Fonn, D., Richter, D., et al. (1997). A patient questionnaire approach
to estimating the prevalence of dry eye symptoms in patients presenting to
optometric practices across Canada. Optometry & Vision Science, 74(8), 624-
631.
Doughty, M. J.,Naase, T. (2008). Nucleus and cell size changes in human bulbar
conjunctival cells after soft contact lens wear, as assessed by impression
cytology. Contact Lens and Anterior Eye, 31(3), 131-140.
du Toit, R., Simpson, T. L., Fonn, D.,Chalmers, R. L. (2001a). Recovery from
hyperemia after overnight wear of low and high transmissibility hydrogel
lenses. Current eye research, 22(1), 68-73.
du Toit, R., Situ, P., Simpson, T.,Fonn, D. (2001b). The effects of six months of
contact lens wear on the tear film, ocular surfaces, and symptoms of
presbyopes. Optom Vis Sci, 78(6), 455-462.
Dua, H. S., Faraj, L. A., Said, D. G., Gray, T.,Lowe, J. (2013). Human corneal
anatomy redefined: a novel pre-Descemet's layer (Dua's layer).
Ophthalmology, 120(9), 1778-1785.
Dutt, R. M., Stocker, E. G., Wolff, C. H., Glavan, I.,Lass, J. H. (1989). A
morphologic and fluorophotometric analysis of the corneal endothelium in
long-term extended wear soft contact lens wearers. Eye & Contact Lens, 15(2),
121-123.
Efron, N. (2007). Contact lens-induced changes in the anterior eye as observed in vivo
with the confocal microscope. Progress in Retinal and Eye Research, 26(4),
398-436.
Efron, N. (2012). Stromal oedema. In N. Efron (Ed.), Contact Lens Complications
(Third Edition) (pp. 185-197). London: W.B. Saunders.
Efron, N., Al-Dossari, M.,Pritchard, N. (2010a). Confocal microscopy of the bulbar
conjunctiva in contact lens wear. Cornea, 29(1), 43-52.
Efron, N., Al‐Dossari, M.,Pritchard, N. (2009). In vivo confocal microscopy of the
bulbar conjunctiva. Clinical & Experimental Ophthalmology, 37(4), 335-344.
Efron, N., Edwards, K., Roper, N., et al. (2010b). Repeatability of measuring corneal
subbasal nerve fiber length in individuals with type 2 diabetes. Eye & contact
lens, 36(5), 245-248.
Bibliography 139
Efron, N., Morgan, P. B., Cameron, I. D., Brennan, N. A.,Goodwin, M. (2007).
Oxygen permeability and water content of silicone hydrogel contact lens
materials. Optometry & Vision Science, 84(4), E328-E337.
Efron, N., Morgan, P. B.,Katsara, S. S. (2001). Validation of grading scales for
contact lens complications. Ophthalmic and Physiological Optics, 21(1), 17-
29.
Efron, N., Perez‐Gomez, I.,Morgan, P. B. (2002). Confocal microscopic observations
of stromal keratocytes during extended contact lens wear. Clinical and
Experimental Optometry, 85(3), 156-160.
Egbert, P., Lauber, S.,Maurice, D. (1977). A simple conjunctival biopsy. American
journal of ophthalmology, 84(6), 798.
Ehlers, N. (1965). The Precorneal Film. Biomicroscopical, Histological and Chemical
Iinvestigations. Acta Ophthalmologica, (Supplement), S81-S134..
Elliott, M., Fandrich, H., Simpson, T.,Fonn, D. (1998). Analysis of the repeatability of
tear break-up time measurement techniques on asymptomatic subjects before,
during and after contact lens wear. Contact Lens and Anterior Eye, 21(4), 98-
103.
Esgin, H.,Erda, N. (2002). Corneal endothelial polymegethism and pleomorphism
induced by daily-wear rigid gas-permeable contact lenses. Eye & Contact
Lens, 28(1), 40-43.
Farris, R. L. (1986). The dry eye: its mechanisms and therapy, with evidence that
contact lens is a cause. Clao j, 12(4), 234-246.
Filippello, M., Cascone, G., Zagami, A.,Scimone, G. (1997). Impression cytology in
Down’s syndrome. British Journal of Ophthalmology, 81(8), 683-685.
Fonn, D., Sweeney, D., Holden, B. A.,Cavanagh, D. (2005). Corneal oxygen
deficiency. Eye & contact lens, 31(1), 23-27.
French, K.,Jones, L. (2008). A decade with silicone hydrogels: Part 2. Optometry
Today, 48(18), 38-42.
Gatlin, J., Melkus, M. W., Padgett, A., et al. (2003). In vivo fluorescent labeling of
corneal wound healing fibroblasts. Experimental Eye Research, 76(3), 361-
371.
Ghinelli, E., Johansson, J., Ríos, J. D., et al. (2003). Presence and localization of
neurotrophins and neurotrophin receptors in rat lacrimal gland. Investigative
Ophthalmology & Visual Science, 44(8), 3352-3357.
Bibliography 140
Gilbard, J. P., Gray, K. L.,Rossi, S. R. (1986). A proposed mechanism for increased
tear-film osmolarity in contact lens wearers. American journal of
ophthalmology, 102(4), 505-507.
Gipson, I. (1994). Anatomy of the conjunctiva, cornea, and limbus. In T. R. Smolin G
(Ed.), The cornea: scientific foundations and clinical practice (3rd ed., pp. 3-
24). Boston: Little, Brown and Company.
Gipson, I. K. (1992). Adhesive mechanisms of the corneal epithelium. Acta
Ophthalmologica, 70(S202), 13-17.
Gipson, I. K. (2004). Distribution of mucins at the ocular surface. Experimental Eye
Research, 78(3), 379-388.
Gipson, I. K. (2016). Goblet cells of the conjunctiva: A review of recent findings.
Progress in Retinal and Eye Research.
Gipson, I. K.,Argüeso, P. (2003). Role of mucins in the function of the corneal and
conjunctival epithelia. International Review of Cytology, 231, 1-49.
Gipson, I. K.,Inatomi, T. (1998). Cellular origin of mucins of the ocular surface tear
film. Advances in Experimental Medicine and Biology, 438, 221-227.
Gipson, I. K., Spurr-Michaud, S., Argüeso, P., et al. (2003). Mucin gene expression in
immortalized human corneal–limbal and conjunctival epithelial cell lines.
Investigative Ophthalmology & Visual Science, 44(6), 2496-2506.
Goto, E.,Tseng, S. C. (2003). Kinetic analysis of tear interference images in aqueous
tear deficiency dry eye before and after punctal occlusion. Investigative
ophthalmology & visual science, 44(5), 1897-1905.
Götz, M., Jaeger, W.,Kruse, F. (1986). Impression cytology as a noninvasive method
of conjunctival biopsy and its results]. Klinische Monatsblätter für
Augenheilkunde, 188(1), 23-28.
Greiner, J. V., Henriquez, A. S., Covington, H. I., Weidman, T. A.,Allansmith, M. R.
(1981). Goblet cells of the human conjunctiva. Archives of Ophthalmology,
99(12), 2190-2197.
Guillon, J. P. (2002). Current clinical techniques to study the tear film and tear
secretions. In D. R. Korb (Ed.), The tear film: structure, function, and clinical
examination (pp. 51-81). Boston US: Butterworth-Heinemann.
Guillon, M.,Maissa, C. (2008). Contact Lens Wear Affects Tear Film Evaporation.
Eye & Contact Lens, 34(6), 326-330.
Guillon, M., Styles, E., Guillon, J. P.,Maissa, C. (1997). Preocular tear film
characteristics of nonwearers and soft contact lens wearers. Optometry &
Vision Science, 74(5), 273-279.
Bibliography 141
Guthoff, R., Baudouin, C.,Stave, J. (2006). Atlas of Confocal Laser Scanning In-vivo
Microscopy in Ophthalmology. Berlin: Springer.
Gwynn, D. R., Stewart, W. C., Hennis, H. L., McMillan, T. A.,Pitts, R. A. (1993).
The influence of age upon inflammatory cell counts and structure of
conjunctiva in chronic open‐angle glaucoma. Acta Ophthalmologica, 71(5),
691-695.
Hamano, H.,Bode, D. (1985). The phenol-red thread test for measuring lacrimation.
Scientific poster, presented at American Academy of Ophthalmology, San
Francisco.
Hamano, H., Maeda, N., Hamano, T., Mitsunaga, S.,Kotani, S. (2008). Corneal
thickness change induced by dozing while wearing hydrogel and silicone
hydrogel lenses. Eye & contact lens, 34(1), 56-60.
Hay, E. M., Thomas, E., Pal, B., et al. (1998). Weak association between subjective
symptoms of and objective testing for dry eyes and dry mouth: results from a
population based study. Annals of the Rheumatic Diseases, 57(1), 20-24.
Haynes, R. J., Tighe, P. J.,Dua, H. S. (1999). Antimicrobial defensin peptides of the
human ocular surface. British Journal of Ophthalmology, 83(6), 737-741.
Henle, F. G. J. (1860). Zur Anatomie der geschlossenen (lenticulären) Drüsen oder
Follikel und der Lymphdrüsen.
Henle, J. (1871). Handbuch der systematischen Anatomie des Menschen (3rd ed.).
Braunschweig: Vieweg.
Hickson, S.,Papas, E. (1997). Prevalence of idiopathic corneal anomalies in a non
contact lens-wearing population. Optometry & Vision Science, 74(5), 293-297.
Hodges, R. R.,Dartt, D. A. (2010). Conjunctival Goblet Cells. In A. D. Darlene (Ed.),
Encyclopedia of the Eye (pp. 369-376). Oxford: Elsevier UK.
Holden, B. A., Reddy, M. K., Sankaridurg, P. R., et al. (1999). Contact lens-induced
peripheral ulcers with extended wear of disposable hydrogel lenses:
histopathologic observations on the nature and type of corneal infiltrate.
Cornea, 18(5), 538-543.
Holden, B. A., Williams, L.,Zantos, S. G. (1985). The etiology of transient endothelial
changes in the human cornea. Investigative Ophthalmology & Visual Science,
26(10), 1354-1359.
Hollingsworth, J. G.,Efron, N. (2004). Confocal microscopy of the corneas of long-
term rigid contact lens wearers. Contact Lens and Anterior Eye, 27(2), 57-64.
Holly, F. (1984). Physical chemistry of the normal and disordered tear film.
Transactions of the Ophthalmological Societies of the United Kingdom,
104(4), 374-380.
Bibliography 142
Holly, F. (2005). The preocular tear film; a small but highly complex part of the eye.
Archivos de la Sociedad Española de Oftalmología, 80(2), 65-68.
Holly, F.,Lemp, M. (1971). Surface chemistry of the tear film: Implications for dry
eye syndromes, contact lenses, and ophthalmic polymers. Journal of Contact
Lens Society of America, 5, 12-19.
Hong, J., Zhu, W., Zhuang, H., et al. (2010). In vivo confocal microscopy of
conjunctival goblet cells in patients with Sjogren's syndrome dry eye. The
British Journal of Ophthalmology, 94(11), 1454-1458.
Hori, Y., Argüeso, P., Spurr-Michaud, S.,Gipson, I. K. (2006). Mucins and contact
lens wear. Cornea, 25(2), 176-181.
Inagaki, Y., Akahori, A., Sugimoto, K., et al. (2003). Comparison of corneal
endothelial bleb formation and disappearance processes between rigid gas-
permeable and soft contact lenses in three classes of dk/l. Eye & Contact Lens,
29(4), 234-237.
Iwata, T., Ohkawa, K.,Uyama, M. (1976). The fine structural localization of
peroxidase activity in goblet cells of the conjunctival epithelium of rats.
Investigative Ophthalmology & Visual Science, 15(1), 40-44.
Jacobsen, I., Jensen, O.,Prause, J. (1984). Structure and composition of Bowman's
membrane. Acta Ophthalmologica, 62(1), 39-53.
Johnson, D. H., Bourne, W. M.,Campbell, R. J. (1982). The ultrastructure of
Descemet's membrane: I. Changes with age in normal corneas. Archives of
Ophthalmology, 100(12), 1942-1947.
Joussen, A. M., Poulaki, V., Mitsiades, N., et al. (2003). VEGF-dependent
conjunctivalization of the corneal surface. Investigative Ophthalmology &
Visual Science, 44(1), 117-123.
Karalezli, A., Borazan, M., Dursun, R., et al. (2011). Impression cytology and ocular
surface characteristics in patients with seborrhoeic dermatitis. Acta
Ophthalmologica, 89(2), e137-e141.
Karalezli, A., Borazan, M., Yilmaz, S., Kiyici, H.,Akova, Y. A. (2009). Conjunctival
impression cytology and tear‐film changes in patients with familial
Mediterranean fever. Acta Ophthalmologica, 87(1), 39-43.
Kawano, K., Uehara, F., Sameshima, M.,Ohba, N. (1984). Application of lectins for
detection of goblet cell carbohydrates of the human conjunctiva. Experimental
Eye Research, 38(5), 439-447.
Keay, L., Jalbert, I., Sweeney, D. F.,Holden, B. A. (2001). Microcysts: clinical
significance and differential diagnosis. Optometry (St. Louis, Mo.), 72(7), 452-
460.
Bibliography 143
Kessing, S. V. (1968). Mucous gland system of the conjunctiva. A quantitative
normal anatomical study. Acta Ophthalmologica Copenhagen, (Suplement),
S1-S95.
Kim, C., Shin, Y. J., Kim, N. J., et al. (2007). Conjunctival epithelial changes induced
by cilia in patients with epiblepharon or entropion. American Journal of
Ophthalmology, 144(4), 564-569.
King-Smith, E., Fink, B., Hill, R., Koelling, K.,Tiffany, J. (2004). The thickness of
the tear film. Current Eye Research, 29(4-5), 357-368.
King-Smith, P. E., Fink, B. A.,Fogt, N. (1999). Three interferometric methods for
measuring the thickness of layers of the tear film. Optometry & Vision
Science, 76(1), 19-32.
Kinoshita, S., Kiorpes, T. C., Friend, J.,Thoft, R. A. (1982). Limbal epithelium in
ocular surface wound healing. Investigative ophthalmology & visual science,
23(1), 73-80.
Knop, E.,Brewitt, H. (1992a). Conjunctival cytology in asymptomatic wearers of soft
contact lenses. Graefe's archive for clinical and experimental ophthalmology,
230(4), 340-347.
Knop, E.,Brewitt, H. (1992b). Induction of conjunctival epithelial alterations by
contact lens wearing. A prospective study. German Journal of Ophthalmology,
1(3-4), 125-134.
Knop, E.,Knop, N. (2010). Conjunctiva Immune Surveillance. In A. D. Darlene (Ed.),
Encyclopedia of the Eye (pp. 356-368). Oxford: Elsevier UK.
Knop, E., Korb, D., Blackie, C.,Knop, N. (2010). The lid margin is an underestimated
structure for preservation of ocular surface health and development of dry eye
disease. Developments in Ophthalmology, 45, 108-122.
Kobayashi, A., Yoshita, T.,Sugiyama, K. (2005). In vivo findings of the
bulbar/palpebral conjunctiva and presumed meibomian glands by laser
scanning confocal microscopy. Cornea, 24(8), 985-988.
Kojima, T., Matsumoto, Y., Dogru, M.,Tsubota, K. (2010). The application of in vivo
laser scanning confocal microscopy as a tool of conjunctival in vivo cytology
in the diagnosis of dry eye ocular surface disease. Molecular Vision, 16, 2457-
2464.
Korb, D. R. (2000). Survey of preferred tests for diagnosis of the tear film and dry
eye. Cornea, 19(4), 483-486.
Korb, D. R. (2002). The tear film: structure, function, and clinical examination:
Elsevier Health Sciences.
Bibliography 144
Korb, D. R., Herman, J. P., Finnemore, V. M., Exford, J. M.,Blackie, C. A. (2008).
An evaluation of the efficacy of fluorescein, rose bengal, lissamine green, and
a new dye mixture for ocular surface staining. Eye & Contact Lens, 34(1), 61-
64.
Krause, W. (1854). About the glands of the conjunctiva. Zeitschr Ration Med, 4, 337-
341.
Kurpakus, M. A., Stock, E. L.,Jones, J. C. (1992). The role of the basement membrane
in differential expression of keratin proteins in epithelial cells. Developmental
Biology, 150(2), 243-255.
Labbé, A., Gheck, L., Iordanidou, V., et al. (2010). An in vivo confocal microscopy
and impression cytology evaluation of pterygium activity. Cornea, 29(4), 392-
399.
Langer, G., Jagla, W., Behrens–Baumann, W., Walter, S.,Hoffmann, W. (1999).
Secretory peptides TFF1 and TFF3 synthesized in human conjunctival goblet
cells. Investigative Ophthalmology & Visual Science, 40(10), 2220-2224.
Le, Q., Wang, W., Hong, J., et al. (2010). An in vivo confocal microscopy and
impression cytology analysis of goblet cells in patients with chemical burns.
Investigative Ophthalmology & Visual Science, 51(3), 1397-1400.
Lemp, M. A.,Gold, J. B. (1986). The effects of extended-wear hydrophilic contact
lenses on the human corneal epithelium. American Journal of Ophthalmology,
101(3), 274-277.
Lievens, C. W., Connor, C. G.,Murphy, H. (2003). Comparing goblet cell densities in
patients wearing disposable hydrogel contact lenses versus silicone hydrogel
contact lenses in an extended-wear modality. Eye & Contact Lens, 29(4), 241-
244.
Linton, R. G., Curnow, D. H.,Riley, W. J. (1961). The meibomian glands: an
investigation into the secretion and some aspects of the physiology. The
British Journal of Ophthalmology, 45(11), 718-723.
Little, S. A.,Bruce, A. S. (1994). Repeatability of the phenol‐red thread and tear
thinning time tests for tear film function. Clinical and Experimental
Optometry, 77(2), 64-68.
Lowther, G. (1997). Examination of patients and predicting tear film-related problems
with hydrogel lens wear. In G. Lowther (Ed.), Dryness, Tears, and Contact
Lens Wear: Clinical Practice in Contact Lenses (pp. 36-77). Newton:
Butterworth-Heinemann.
MacRae, S. M., Matsuda, M.,Phillips, D. S. (1994). The long-term effects of
polymethylmethacrylate contact lens wear on the corneal endothelium.
Ophthalmology, 101(2), 365-370.
Bibliography 145
Maruyama, K., Yokoi, N., Takamata, A.,Kinoshita, S. (2004). Effect of environmental
conditions on tear dynamics in soft contact lens wearers. Investigative
ophthalmology & visual science, 45(8), 2563-2568.
Mastropasqua, L., Agnifili, L., Fasanella, V., et al. (2013). Conjunctival goblet cells
density and preservative‐free tafluprost therapy for glaucoma: an in vivo
confocal microscopy and impression cytology study. Acta Ophthalmologica,
91(5), 397-405.
Matsumoto, Y., Dogru, M., Goto, E., et al. (2008). Alterations of the tear film and
ocular surface health in chronic smokers. Eye, 22(7), 961-968.
Maurice, D. (1970). The transparency of the corneal stroma. Vision Research, 10(1),
107-108.
McCarty, C. A., Bansal, A. K., Livingston, P. M., Stanislavsky, Y. L.,Taylor, H. R.
(1998). The epidemiology of dry eye in Melbourne, Australia. Ophthalmology,
105(6), 1114-1119.
McMahon, T. T., Polse, K. A., McNamara, N. Y.,Viana, M. A. (1996). Recovery
from induced corneal edema and endothelial morphology after long-term
PMMA contact lens wear. Optometry & Vision Science, 73(3), 184-188.
McMonnies, C. (1986). Marginal dry eye diagnosis: history versus biomicroscopy. In
F. J. Holly (Ed.), The Pre-Ocular Tear Film in Health, Disease and Contact
Lens Wear (pp. 32-40). Lubbock, TX: Dry Eye Institute.
McMonnies, C.,Ho, A. (1987). Patient history in screening for dry eye conditions.
Journal of the American Optometric Association, 58(4), 296-301.
McNamara, N. A., Van, R., Tuchin, O. S.,Fleiszig, S. M. (1999). Ocular surface
epithelia express mRNA for human beta defensin-2. Experimental Eye
Research, 69(5), 483-490.
Mengher, L. S., Bron, A. J., Tonge, S. R.,Gilbert, D. J. (1985). A non-invasive
instrument for clinical assessment of the pre-corneal tear film stability.
Current Eye Research, 4(1), 1-7.
Messmer, E. M. (2008). In Vivo Confocal Microscopy in Healthy Conjunctiva,
Conjunctivitis, and Conjunctival Tumors. In T. Reinhard (Ed.), Cornea and
External Eye Disease (pp. 217-227). Berlin: Springer.
Messmer, E. M., Mackert, M. J., Zapp, D. M.,Kampik, A. (2006). In vivo confocal
microscopy of normal conjunctiva and conjunctivitis. Cornea, 25(7), 781-788.
Metz, A. (1868). The Conjunctiva. In A. Metz (Ed.), The anatomy and histology of
the human eye (pp. 165-175). Philadelphia: Medical and surgical reporter.
Bibliography 146
Miyashita, K., Azuma, N.,Hida, T. (1992). Morphological and histochemical studies
of goblet cells in developing human conjunctiva. Japanese Journal of
Ophthalmology, 36(2), 169-174.
Mobilia, E. F.,Foster, C. S. (1978). The Management of Recurrent Corneal Erosions
with Ultra-Thin Lenses. Eye & Contact Lens, 4(1), 25-29.
Moon, J. W., Shin, K. C., Lee, H. J., et al. (2006). The effect of contact lens wear on
the ocular surface changes in keratoconus. Eye & Contact Lens, 32(2), 96-101.
Morales-Fernández, L., Pérez-Álvarez, J., García-Catalán, R., Benítez-del-Castillo,
J.,García-Sánchez, J. (2010). Clinical-histological correlation in patients with
dry eye. Archivos de la Sociedad Española de Oftalmología (English Edition),
85(7), 239-245.
Moreno, M., Villena, A., Cabarga, C., Sanchez-Font, E.,Garcia-Campos, J. (2003).
Impression cytology of the conjunctival epithelium after antiglaucomatous
treatment with latanoprost. European journal of Ophthalmology, 13(6), 553-
559.
Mrugacz, M., Kasacka, I., Bakunowicz-Lazarczyk, A., Kaczmarski, M.,Kulak, W.
(2008). Impression cytology of the conjunctival epithelial cells in patients with
cystic fibrosis. Eye, 22(9), 1137-1140.
Murube, J.,Rivas, L. (2003). Impression cytology on conjunctiva and cornea in dry
eye patients establishes a correlation between squamous metaplasia and dry
eye clinical severity. European Journal of Ophthalmology, 13(2), 115-127.
Natadisastra, G., Wittpenn, J. R., Muhilal, et al. (1988). Impression cytology: a
practical index of vitamin A status. [Article]. American Journal of Clinical
Nutrition, 48(3), 695-701.
Nelson, J. (1995). Simultaneous evaluation of tear turnover and corneal epithelial
permeability by fluorophotometry in normal subjects and patients with
keratoconjunctivitis sicca (KCS). Transactions of the American
Ophthalmological Society, 93, 709-753.
Nelson, J.,Wright, J. (1986). Impression cytology of the ocular surface in
keratoconjunctivitis sicca. The preocular tear film in health, disease, and
contact lens wear. Lubbock: Texas Dry Eye Institute, 140-156.
Nelson, J. D. (1988). Impression cytology. Cornea, 7(1), 71-81.
Nelson, J. D.,Wright, J. C. (1984). Conjunctival goblet cell densities in ocular surface
disease. Archives of Ophthalmology, 102(7), 1049-1051.
Nichols, J. J. (2015). 2014 Annual Report. Contact Lens Spectrum, 30(1), 22-27.
Bibliography 147
Nichols, J. J., Nichols, K. K., Puent, B., Saracino, M.,Mitchell, G. L. (2002).
Evaluation of tear film interference patterns and measures of tear break-up
time. Optometry & Vision Science, 79(6), 363-369.
Nichols, J. J.,Sinnott, L. T. (2006). Tear film, contact lens, and patient-related factors
associated with contact lens–related dry eye. Investigative ophthalmology &
visual science, 47(4), 1319-1328.
Nichols, J. J., Ziegler, C., Mitchell, G. L.,Nichols, K. K. (2005). Self-Reported Dry
Eye Disease across Refractive Modalities. Investigative ophthalmology &
visual science, 46(6), 1911-1914.
Nichols, K. K., Redfern, R. L., Jacob, J. T., et al. (2013). The TFOS International
Workshop on Contact Lens Discomfort: Report of the Definition and
Classification SubcommitteeReport of the Definition and Classification
Subcommittee. Investigative Ophthalmology & Visual Science, 54(11),
TFOS14-TFOS19.
Nieuwendaal, C., Odenthal, M., Kok, J., et al. (1994). Morphology and function of the
corneal endothelium after long-term contact lens wear. Investigative
Ophthalmology & Visual Science, 35(7), 3071-3077.
Nishida, K., Kinoshita, S., Ohashi, Y., Kuwayama, Y.,Yamamoto, S. (1995). Ocular
surface abnormalities in aniridia. American Journal of Ophthalmology, 120(3),
368-375.
Norn, M. (1969). Dead, degenerate, and living cells in conjunctival fluid and mucous
thread. Acta Ophthalmologica, 47(5‐6), 1102-1115.
Orsborn, G.,Robboy, M. (1989). Hydrogel lenses and dry-eye symptoms. Journal of
the British Contact Lens Association, 12, 37-38.
Paschides, C., Petroutsos, G.,Psilas, K. (1991). Correlation of conjunctival impression
cytology results with lacrimal function and age. Acta Ophthalmologica, 69(4),
422-425.
Patel, S., Murray, D., McKenzie, A., Shearer, D.,McGrath, B. (1985). Effects of
fluorescein on tear breakup time and on tear thinning time. American Journal
of Optometry and Physiological Optics, 62(3), 188-190.
Patel, S. V., McLaren, J. W., Hodge, D. O.,Bourne, W. M. (2001). Normal human
keratocyte density and corneal thickness measurement by using confocal
microscopy in vivo. Investigative Ophthalmology & Visual Science, 42(2),
333-339.
Pflugfelder, S. C., Tseng, S. C., Sanabria, O., et al. (1998). Evaluation of subjective
assessments and objective diagnostic tests for diagnosing tear-film disorders
known to cause ocular irritation. Cornea, 17(1), 38-56.
Bibliography 148
Pimenides, D., Steele, C., McGhee, C.,Bryce, I. (1996). Deep corneal stromal
opacities associated with long term contact lens wear. British Journal of
Ophthalmology, 80(1), 21-24.
Pisella, P. J., Brignole, F., Debbasch, C., et al. (2000). Flow cytometric analysis of
conjunctival epithelium in ocular rosacea and keratoconjunctivitis sicca.
Ophthalmology, 107(10), 1841-1849.
Pisella, P. J., Malet, F., Lejeune, S., et al. (2001). Ocular surface changes induced by
contact lens wear. Cornea, 20(8), 820-825.
Popper, M., nika, o., Morgado, A., et al. (2003). Corneal cell density measurement in
vivo by scanning slit confocal microscopy: method and validation. Ophthalmic
Research, 36(5), 270-276.
Prabhasawat, P.,Tseng, S. C. G. (1997). Impression cytology study of epithelial
phenotype of ocular surface reconstructed by preserved human amniotic
membrane. Archives of ophthalmology, 115(11), 1360-1367.
Price-Schiavi, S., Meller, D., Jing, X., et al. (1998). Sialomucin complex at the rat
ocular surface: a new model for ocular surface protection. Biochem, 335, 457-
463.
Pritchard, N., Fonn, D.,Brazeau, D. (1999). Discontinuation of contact lens wear: a
survey. International Contact Lens Clinic, 26(6), 157-162.
Rajagopalan, K., Abetz, L., Mertzanis, P., et al. (2005). Comparing the discriminative
validity of two generic and one disease-specific health-related quality of life
measures in a sample of patients with dry eye. Value in Health, 8(2), 168-174.
Rath, R., Stave, J., Guthoff, R., Giebel, J.,Tost, F. (2006). [In vivo imaging of the
conjunctival epithelium using confocal laser scanning microscopy]. Der
Ophthalmologe: Zeitschrift der Deutschen Ophthalmologischen Gesellschaft,
103(5), 401-405.
Reinstein, D. Z., Archer, T. J., Gobbe, M., Silverman, R. H.,Coleman, D. J. (2008).
Epithelial thickness in the normal cornea: three-dimensional display with
Artemis very high-frequency digital ultrasound. Journal of Refractive Surgery,
24(6), 571-581.
Report of the International Dry Eye WorkShop (DEWS). (2007). The Ocular Surface,
5(2), 65-204
doi:10.1016/s1542-0124(12)70080-0
Ridley, H. (1963). Embryology of the Eye. British Medical Journal, 2(5369), 1398-
1399.
Riley, C., Young, G.,Chalmers, R. (2006). Prevalence of ocular surface symptoms,
signs, and uncomfortable hours of wear in contact lens wearers: the effect of
Bibliography 149
refitting with daily-wear silicone hydrogel lenses (senofilcon a). Eye &
contact lens, 32(6), 281-286.
Rios, J. D., Zoukhri, D., Rawe, I. M., et al. (1999). Immunolocalization of muscarinic
and VIP receptor subtypes and their role in stimulating goblet cell secretion.
Investigative Ophthalmology & Visual Science, 40(6), 1102-1111.
Rivas, L., Alvarez, M., Rodriguez, J.,Murube, J. (1995). Ophthalmological tests in
patients with keratoconjunctivitis sicca with and without association of
primary Sjogren's syndrome. German Journal of Ophthalmology, 4(5), 306-
310.
Rivas, L., Oroza, M. A., Perez‐Esteban, A.,Murube‐del‐Castillo, J. (1991).
Topographical distribution of ocular surface cells by the use of impression
cytology. Acta Ophthalmologica, 69(3), 371-376.
Rivas, L., Rodriguez, J. J., Alvarez, M. I., Oroza, M. A.,Castillo, J. M. (1993).
Correlation between impression cytology and tear function parameters in
Sjögren's syndrome. Acta Ophthalmologica, 71(3), 353-359.
Rodriguez-Prats, J. L., Hamdi, I. M., Rodriguez, A. E., Galal, A.,Alio, J. L. (2007).
Effect of suction ring application during LASIK on goblet cell density.
Journal of Refractive Surgery, 23(6), 559-562.
Rodriguez, A. E., Rodriguez-Prats, J. L., Hamdi, I. M., et al. (2007). Comparison of
goblet cell density after femtosecond laser and mechanical microkeratome in
LASIK. Investigative Ophthalmology & Visual Science, 48(6), 2570-2575.
Rojas, M., Rodriguez, M., Blanco, J. C.,Salorio, M. (1993). Impression cytology in
patients with keratoconjunctivitis sicca. Cytopathology, 4(6), 347-355.
Rolando, M., Brezzo, V.,Calabria, G. (1994). Ocular surface changes induced by
repeated impression cytology. Advances in Experimental Medicine and
Biology, 350, 249-254.
Rolando, M., Terragna, F., Giordano, G.,Calabria, G. (1990). Conjunctival Surface
Damage Distribution in Keratoconjunctivitis sicca An Impression Cytology
Study. Ophthalmologica, 200(4), 170-176.
Rolando, M.,Zierhut, M. (2001). The ocular surface and tear film and their
dysfunction in dry eye disease. Survey of Ophthalmology, 45 (Supplement),
S203-S210.
Rummenie, V. T., Matsumoto, Y., Dogru, M., et al. (2008). Tear cytokine and ocular
surface alterations following brief passive cigarette smoke exposure. Cytokine,
43(2), 200-208.
Saini, J., Rajwanshi, A.,Dhar, S. (1990). Clinicopathological correlation of hard
contact lens related changes in tarsal conjunctiva by impression cytology. Acta
Ophthalmologica, 68(1), 65-70.
Bibliography 150
Satici, A., Bitiren, M., Ozardali, I., et al. (2003). The effects of chronic smoking on
the ocular surface and tear characteristics: a clinical, histological and
biochemical study. Acta Ophthalmologica Scandinavica, 81(6), 583-587.
Schaumberg, D. A., Sullivan, D. A., Buring, J. E.,Dana, M. R. (2003). Prevalence of
dry eye syndrome among US women. American Journal of Ophthalmology,
136(2), 318-326.
Schein, O. D., Hochberg, M. C., Munoz, B., et al. (1999). Dry eye and dry mouth in
the elderly: a population-based assessment. Archives of Internal Medicine,
159(12), 1359-1363.
Schein, O. D., Munoz, B., Tielsch, J. M., Bandeen-Roche, K.,West, S. (1997).
Prevalence of dry eye among the elderly. American Journal of
Ophthalmology, 124(6), 723-728.
Schiffman, R. M., Christianson, M. D., Jacobsen, G., Hirsch, J. D.,Reis, B. L. (2000).
Reliability and validity of the ocular surface disease index. Archives of
Ophthalmology, 118(5), 615-621.
Schirmer, O. (1903). Studien zur physiologie und pathologie der Tränenabsonderung
und Tränenabfuhr. Graefe's Archive for Clinical and Experimental
Ophthalmology, 56(2), 197-291.
Scott, J. E. (1988). Proteoglycan-fibrillar collagen interactions. Biochemical Journal,
252(2), 313-323.
Sellheyer, K.,Spitznas, M. (1988). Ultrastructural observations on the development of
the human conjunctival epithelium. Graefe's Archive for Clinical and
Experimental Ophthalmology, 226(5), 489-499.
Şengör, T., Gürdal, C., Kirimlioglu, H., İrkeç, M.,Aydin, S. (2002). Colour-coded
mapping technique in impression cytology–Findings in soft contact lens
wearers and patients with other external eye diseases. Ophthalmologica,
216(3), 155-158.
Setälä, K., Vasara, K., Vesti, E.,Ruusuvaara, P. (1998). Effects of long‐term contact
lens wear on the corneal endothelium. Acta Ophthalmologica Scandinavica,
76(3), 299-303.
Shatos, M. A., Rios, J. D., Horikawa, Y., et al. (2003). Isolation and characterization
of cultured human conjunctival goblet cells. Investigative Ophthalmology &
Visual Science, 44(6), 2477-2486.
Sheng, H.,Bullimore, M. A. (2007). Factors affecting corneal endothelial morphology.
Cornea, 26(5), 520-525.
Bibliography 151
Simon, P., Jaison, S. G., Chopra, S. K.,Jacob, S. (2002). Conjunctival impression
cytology in contact lens wearers. Indian Journal of Ophthalmology, 50(4),
301-306.
Smith, J. A. (2007). The epidemiology of dry eye disease: report of the Epidemiology
Subcommittee of the International Dry Eye WorkShop (2007). The Ocular
Surface, 5(2), 93-107.
Solomon, A., Kaiserman, I., Raiskup, F. D., Landau, D.,Frucht-Pery, J. (2004). Long-
term effects of mitomycin C in pterygium surgery on scleral thickness and the
conjunctival epithelium. Ophthalmology, 111(8), 1522-1527.
Stern, M. E., Beuerman, R. W., Fox, R. I., et al. (1998). The pathology of dry eye: the
interaction between the ocular surface and lacrimal glands. Cornea, 17(6),
584-589.
Stretton, S., Jalbert, I.,Sweeney, D. F. (2003). Corneal hypoxia secondary to contact
lenses: the effect of high-Dk lenses. Ophthalmology clinics of North America,
16(3), 327-340, v.
Sullivan, B. D., Crews, L. A., Sönmez, B., et al. (2012). Clinical utility of objective
tests for dry eye disease: variability over time and implications for clinical
trials and disease management. Cornea, 31(9), 1000-1008.
Tao, A., Wang, J., Chen, Q., et al. (2011). Topographic thickness of Bowman's layer
determined by ultra-high resolution spectral domain–optical coherence
tomography. Investigative Ophthalmology & Visual Science, 52(6), 3901-
3907.
Taylor, C. R. (2015). Quantitative in situ proteomics; a proposed pathway for
quantification of immunohistochemistry at the light-microscopic level. Cell
and Tissue Research, 360(1), 109-120.
Tiffany, J. M.,Bron, A. J. (1978). Role of tears in maintaining corneal integrity.
Transactions of the Ophthalmological Societies of the United Kingdom, 98(3),
335-338.
Tighe, B. J. (2006). Contact Lens Materials. In A. Phillips & L. Speedwell (Eds.),
Contact Lenses (pp. 59-78). Edinburgh.
Tomatir, D. K., Erda, N.,Gürlü, V. P. (2008). Effects of different contact lens
materials and contact lens-wearing periods on conjunctival cytology in
asymptomatic contact lens wearers. Eye & Contact Lens, 34(3), 166-168.
Tomlinson, A., Blades, K. J.,Pearce, E. I. (2001). What does the phenol red thread test
actually measure? Optometry & Vision Science, 78(3), 142-146.
Tomlinson, A., Khanal, S., Ramaesh, K., Diaper, C.,McFadyen, A. (2006). Tear film
osmolarity: determination of a referent for dry eye diagnosis. Investigative
ophthalmology & visual science, 47(10), 4309-4315.
Bibliography 152
Tseng, S.-H., Chen, Y.-T., Cheng, H.-C., et al. (2001). Impression cytology study of
conjunctival epithelial phenotypes on the healing ocular surface after
pterygium excision. Cornea, 20(3), 244-250.
Tseng, S. (1985). Staging of conjunctival squamous metaplasia by impression
cytology. Ophthalmology, 92(6), 728-733.
Tseng, S., Hirst, L., Farazdaghi, M.,Green, W. (1984). Goblet cell density and
vascularization during conjunctival transdifferentiation. Investigative
Ophthalmology & Visual Science, 25(10), 1168-1176.
van Es, J. H., van Gijn, M. E., Riccio, O., et al. (2005). Notch/γ-secretase inhibition
turns proliferative cells in intestinal crypts and adenomas into goblet cells.
Nature, 435(7044), 959-963.
Van Haeringen, N. J. (1981). Clinical biochemistry of tears. Survey of
Ophthalmology, 26(2), 84-96.
Vannas, A., Holden, B. A.,Makittie, J. (1984). The ultrastructure of contact lens
induced changes. Acta Ophthalmologica, 62(2), 320-333.
Villani, E., Beretta, S., Galimberti, D., Viola, F.,Ratiglia, R. (2011a). In Vivo
Confocal Microscopy of Conjunctival Roundish Bright Objects: Young,
Older, and Sjögren Subjects. Investigative Ophthalmology & Visual Science,
52(7), 4829-4832.
Villani, E., Ceresara, G., Beretta, S., et al. (2011b). In Vivo Confocal Microscopy of
Meibomian Glands in Contact Lens Wearers. Investigative Ophthalmology &
Visual Science, 52(8), 5215-5219.
Vitali, C., Bombardieri, S., Jonsson, R., et al. (2002). Classification criteria for
Sjögren's syndrome: a revised version of the European criteria proposed by the
American-European Consensus Group. Annals of the Rheumatic Diseases,
61(6), 554-558.
Vujković, V., Mikac, G.,Kozomara, R. (2001). Distribution and density of
conjunctival goblet cells. Medicinski Pregled, 55(5-6), 195-200.
Wakamatsu, T. H., Okada, N., Kojima, T., et al. (2009). Evaluation of conjunctival
inflammatory status by confocal scanning laser microscopy and conjunctival
brush cytology in patients with atopic keratoconjunctivitis (AKC). Molecular
Vision, 15, 1611-1619.
Wang, Y., Dogru, M., Matsumoto, Y., et al. (2007). The Impact of Nasal
Conjunctivochalasis on Tear Functions and Ocular Surface Findings.
American Journal of Ophthalmology, 144(6), 930-937.
Wang, Y., Ogawa, Y., Dogru, M., et al. (2010). Baseline profiles of ocular surface
and tear dynamics after allogeneic hematopoietic stem cell transplantation in
Bibliography 153
patients with or without chronic GVHD-related dry eye. Bone Marrow
Transplant, 45(6), 1077-1083.
Waring, G. O. (1982). Posterior collagenous layer of the cornea: ultrastructural
classification of abnormal collagenous tissue posterior to Descemet's
membrane in 30 cases. Archives of Ophthalmology, 100(1), 122-134.
Watanabe, H., Ohashi, Y., Kinoshita, S., Manabe, R.,Ohshiden, K. (1993).
Distribution of epidermal growth factor in rat ocular and periocular tissues.
Graefe's Archive for Clinical and Experimental Ophthalmology, 231(4), 228-
232.
Wei, A., Hong, J., Sun, X.,Xu, J. (2011). Evaluation of age-related changes in human
palpebral conjunctiva and meibomian glands by in vivo confocal microscopy.
Cornea, 30(9), 1007-1012.
Wiffen, S. J., Hodge, D. O.,Bourne, W. M. (2000). The effect of contact lens wear on
the central and peripheral corneal endothelium. Cornea, 19(1), 47-51.
Wolff, E. (1946). The mucocutaneous junction of the lidmargin and the distribution of
the tear fluid. Transactions of the American Ophthalmological Society, 66,
291-308.
Yamabayashi, S.,Tsukahara, S. (1987). Histochemical studies on the conjunctival
goblet cells. Ophthalmic Research, 19(3), 137-140.
Yasueda, S., Yamakawa, K., Nakanishi, Y., Kinoshita, M.,Kakehi, K. (2005).
Decreased mucin concentrations in tear fluids of contact lens wearers. Journal
of Pharmaceutical and Biomedical Analysis, 39(1), 187-195.
Yoon, K.-C., Song, B.-Y.,Seo, M.-S. (2005). Effects of smoking on tear film and
ocular surface. Korean Journal of Ophthalmology, 19(1), 18-22.
Yoon, K. C., Im, S. K.,Seo, M. S. (2004). Changes of tear film and ocular surface in
diabetes mellitus. Korean Journal of Ophthalmology, 18(2), 168-174.
Zantos, S.,Holden, B. (1977). Transient endothelial changes soon after wearing soft
contact lenses. American Journal of Optometry and Physiological Optics,
54(12), 856-858.
Zhang, B.,Yao, Y. (2002). [Application of impression cytology in diagnosis of ocular
surface diseases]. Zhejiang da xue xue bao. Yi xue ban= Journal of Zhejiang
University. Medical sciences, 31(5), 383-387.
Zhang, X., Li, Q., Liu, B., et al. (2011). In vivo cross-sectional observation and
thickness measurement of bulbar conjunctiva using optical coherence
tomography. Investigative ophthalmology & visual science, 52(10), 7787-
7791.
Bibliography 154
Zhang, Y.-g., Wu, S., Lu, R., et al. (2015). Tight junction CLDN2 gene is a direct
target of the vitamin D receptor. Scientific Reports, 5, 10642.
Zhivov, A., Beck, R.,Guthoff, R. F. (2009). Corneal and conjunctival findings after
mitomycin C application in pterygium surgery: an in-vivo confocal
microscopy study. Acta Ophthalmologica, 87(2), 166-172.
Zhivov, A., Stachs, O., Kraak, R., Stave, J.,Guthoff, R. F. (2006). In vivo confocal
microscopy of the ocular surface. The Ocular Surface, 4(2), 81-93.
Zhu, W., Hong, J., Zheng, T., et al. (2010). Age-related changes of human conjunctiva
on in vivo confocal microscopy. British Journal of Ophthalmology, 94(11),
1448-1453.
Zhu, W. Q., Xu, J. J., Sun, X. H., Zheng, T. Y.,Le, Q. H. (2009). [Normal human
bulbar conjunctiva on confocal microscopy in vivo]. Zhonghua Yan Ke Za Zhi,
45(4), 344-349.
Appendix A: Ethics clearance and consent form 155
Appendices
Appendix A: Ethics Clearance and Consent Form
Appendix A: Ethics clearance and consent form 156
Appendix A: Ethics clearance and consent form 157
Appendix A: Ethics clearance and consent form 158
Appendix A: Ethics clearance and consent form 159
PARTICIPANT INFORMATION FOR QUT RESEARCH PROJECT
Impact of contact lens wear on conjunctival goblet and Langerhans cells
QUT Ethics Approval Number 1300000117
RESEARCH TEAM
Principal Researchers:
Ms Luisa Holguin, Masters student and Mr Yahya Alzahrani, PhD student
Associate Researchers:
Professor Nathan Efron and Dr Nicola Pritchard, Queensland University of Technology (QUT)
DESCRIPTION
This project is being undertaken as part of Masters research for student Luisa Holguin and PhD research for Yahya Alzahrani.
The study will explore two types of cells that occur naturally in both the cornea and conjunctiva. The purpose of this project is to examine the characteristics of these cells in the cornea the front clear window of the eye) and conjunctiva (the glistening clear tissue seen covering the white of the eye) and to identify any potential consequences of contact lens wear. It is possible that the mucus-producing cells in the conjunctiva called goblet cells increase or decrease in number when contact lenses are worn. It is also possible that Langerhans cells (cells recruited into the tissue when the eye is inflamed) are altered in number or appearance in contact lens wear. It is hoped the outcome of these studies will provide new information that may help contact lens wearers and provide eye care practitioners with advice regarding care of their patients.
Goblet in the conjunctiva and Langerhans cells in the cornea and conjunctiva play an important role in keeping the eye healthy. Goblet cells have an important role in maintaining wettability of the eye surface and Langerhans cells have a protective and healing role in the eye. The impact of contact lenses on the eye have been studied for many years, however, new technology continues to permit further exploration and understanding of the clinical effects of contact lens wear. The technique that will be used to measure the goblet and Langerhans cells is called confocal microscopy. This is a non-invasive test for the examination of eye tissue at approximately 400x magnification. Another technique to examine goblet cells is called impression cytology, where a sample of surface cells is taken from the eye by touching the surface of the conjunctiva with a sterile piece of acetate paper. The cells are then stained and counted using a light microscope.
You are invited to participate in this project because you are aged between 18 and 70 and have not worn contact lenses (or not worn contact lenses for at least 6 months) and may be interested in the use of contact lens for correction of vision.
PARTICIPATION
Your participation will involve:
Answering questions about your eye and medical history;
An examination of the front part of your eye using microscope at about 40x magnification – this takes about 3 minutes;
Sitting an instrument, having a drop of anaesthetic placed in both eyes (to numb the eye) and having images captured of the superficial layers of cells of the eye – this takes about 5 minutes. The cells will be counted from these images;
Also, while seated, an “impression” of cells will be taken from a 10 mm region on the least sensitive part of the eye – the nasal conjunctiva – this takes about 2 minutes. The cells will be
Appendix A: Ethics clearance and consent form 160
stained and counted under a light microscope.
4 visits to the Anterior Eye Laboratory at the Institute of Health and Biomedical Innovation (IHBI) at QUT Kelvin Grove – at baseline, 1 week, 1 month and 6 months. The first baseline visit will be about 2 hours long; the remaining visits will be approximately 30 minutes long.
You will be asked not to rub your eyes for at least 40 minutes after the tests because the drop numbs the eye, and it is possible for you to damage the front layer of cells of your eye without noticing it.
Your participation in this project is entirely voluntary. If you do agree to participate you can withdraw from the project without comment or penalty. If you withdraw, on request any identifiable information already obtained from you will be destroyed. Your decision to participate or not participate will in no way impact upon your current or future relationship with QUT (for example your grades).
EXPECTED BENEFITS
If you are able to be fitted with soft contact lenses, you will receive these at no cost to you for the duration of the study. There is no anticipated benefit to those who will not wear contact lenses in the study (the control group), however, the knowledge gained from the study maybe benefit people who wear contact lenses in the future. Some people find participating in studies an interesting experience.
RISKS
There is no risk beyond that involved in a regular eye examination. During and at the end of the study the front of your eyes will be examined again. If the investigator believes it is in your best interests, you may be asked to return for a follow-up examination in addition to the set schedule to check the health of your eyes. The study does not replace full eye care because this study only involves the front part of the eye.
PRIVACY AND CONFIDENTIALITY
All comments and responses will be treated confidentially, and if presented or published, you will be anonymous.
CONSENT TO PARTICIPATE
We would like to ask you to sign a written consent form (enclosed) to confirm your agreement to participate.
QUESTIONS / FURTHER INFORMATION ABOUT THE PROJECT
If have any questions or require any further information please contact one of the research team members below.
Ms Luisa Holguin – Principal Researcher Dr Nicola Pritchard – Supervisor
School of Optometry and Vision Science and IHBI
School of Optometry and Vision Science and IHBI
31386404 [email protected] 3138 6414 [email protected]
CONCERNS / COMPLAINTS REGARDING THE CONDUCT OF THE PROJECT
QUT is committed to research integrity and the ethical conduct of research projects. However, if you do have any concerns or complaints about the ethical conduct of the project you may contact the QUT Research Ethics Unit on 3138 5123 or email [email protected]. The QUT Research Ethics Unit is not connected with the research project and can facilitate a resolution to your concern in an impartial manner.
Thank you for helping with this research project. Please keep this sheet for your
information.
Appendix A: Ethics clearance and consent form 161
CONSENT FORM FOR QUT RESEARCH PROJECT
Impact of contact lens wear on conjunctival goblet and Langerhans cells
QUT Ethics Approval Number 1300000117
RESEARCH TEAM CONTACTS
Ms Luisa Holguin Mr Yahya Alzahrani 3138 6404 0421 808 117 [email protected]
Prof Nathan Efron Dr Nicola Pritchard 3138 6401 3138 6414 [email protected] [email protected]
STATEMENT OF CONSENT
By signing below, you are indicating that you:
Have read and understood the information document regarding this project.
Have had any questions answered to your satisfaction.
Understand that if you have any additional questions you can contact the research team.
Understand that you are free to withdraw at any time, without comment or penalty.
Understand that you can contact the Research Ethics Unit on 3138 5123 or email [email protected] if you have concerns about the ethical conduct of the project.
Agree to participate in the project.
Name
Signature
Date
Please return this sheet to the investigator. You will be given a copy of the document
to
Appendix B Conferences, presentations, publications and awards 163
Appendix B: Conference Presentations, Publications and Awards Arising from this Thesis
1. Colorado, Luisa H.; Alzahrani, Yahya; Pritchard, Nicola; Efron, Nathan.
Impact of Contact Lens Wear on Conjunctival Goblet Cells. The Association
for Research in Vision and Ophthalmology (ARVO), 2015, Denver, Colorado,
USA. (Poster presentation)
2. Colorado, Luisa H.; Alzahrani, Yahya; Pritchard, Nicola; Efron, Nathan.
Impact of Contact Lens Wear on Conjunctival Goblet Cells. International
Cornea and Contact Lens Congress (ICCLC), 2015, Gold Coast, Queensland,
Australia. (Oral presentation)
3. Colorado, Luisa H.; Alzahrani, Yahya; Pritchard, Nicola; Efron, Nathan.
Longitudinal assessment of conjunctival goblet cell density in contact lens-
associated dry eye. International Society of Contact Lens Research (ISCLR),
2015, Budapest, Hungary. (Oral and poster presentation)
4. Cornea and Contact Lens Society Research Award 2014.
5. Cornea and Contact Lens Society Research Award 2015.
6. Colorado LH, Alzahrani Y, Pritchard N, Efron N. Assessment of conjunctival
goblet cell density using laser scanning confocal microscopy versus
impression cytology. Contact Lens Ant Eye 2016; 39(3): 221-226.
(Publication)
7. Colorado LH, Pritchard N, Cronin BG, Efron N. Characterization of goblet
cells in a pterygium biopsy using laser scanning confocal microscopy and
immunohistochemistry. Cornea 2016; 35(8): 1127-1131. (Publication)
8. Colorado LH, Alzahrani Y, Pritchard N, Efron N. Time course of changes in
goblet cell density in symptomatic and asymptomatic contact lens wearers.
Invest Ophthalmol Vis Sci 2016; 57(6): 2560-2566. (Publication)