Epithelial Immune Cell Response to Initial Soft Contact Lens Wear in the Human Corneal and Conjunctival Epithelium

Abstract

Purpose: The purpose of this study was to assess the immediate ocular immune response to soft contact lens (CL) wear by examining presumed epithelial immune cell (EIC) density and morphology at the central, peripheral, limbal cornea, and conjunctiva.

Methods: Fifty-four participants naïve to CL wear (mean age = 24.8 ± 9.8 years, 44% female participants), were examined using in vivo confocal microscopy at baseline and after 2 hours of CL wear (1-Day ACUVUE MOIST). Images were captured at the central, temporal far peripheral and limbal cornea, and bulbar conjunctiva. EIC density was counted manually and morphology was graded. Differences in EIC parameters pre- and post-CL wear were examined using a generalized estimating equation model with appropriate post hoc analyses.

Results: After 2 hours of soft CL wear, there was a significant increase in EIC density in all regions other than the central cornea (all P < 0.001). Cell body size was significantly larger, and a higher proportion of participants exhibited EIC with long dendrites after lens wear at the central and peripheral cornea (both P < 0.001). There was a significant increase in the number of participants displaying EIC with thick dendrites at the peripheral (P = 0.04) and limbal cornea (P < 0.001) after lens wear.

Conclusions: EICs were primarily recruited to the peripheral regions, whereas the central cornea shows no significant recruitment after short-term CL wear. Both central and peripheral corneas exhibited an enhanced antigen capture capacity, whereas migratory capacity was increased in the peripheral corneal regions suggesting EIC activation following a short period of CL wear.

Contact lenses (CLs) are widely used as a means of correcting refractive errors and for cosmetic purposes to change the color of the eyes.¹ Whereas CLs are relatively safe, the introduction of a foreign material to the ocular surface carries a risk of inflammatory responses and CL wear remains a significant contributing factor to corneal infiltrative events.^2–5 Improving the current understanding of the ocular immune response during CL wear through the observation of corneal and conjunctival immune cells, such as dendritic cells, may help to address this problem.^6–8 Dendritic cells are specialized antigen-presenting cells that perform a role as sentinels of the body's immunity and are responsible for modulating the innate and adaptive immune systems.⁹ 

In healthy human corneal epithelium, cells presumed to be immune cells are identified using in vivo confocal microscopy (IVCM), with density decreasing from the limbus toward the central cornea.^10–14 However, in the presence of corneal inflammation and infection, the density of these cells increases at almost all ocular surface regions, including the central, peripheral, and limbal cornea, and conjunctiva.¹⁵ These cells also undergo alterations in morphology,^13,16,17 these morphology alterations include changes in cell body size and in the presence and appearance of dendrites.¹⁸ 

The density of epithelial immune cells (EICs) in the cornea is also elevated during uncomplicated long-term soft CL wear, with accompanying morphological changes, suggesting an enhanced immune response.^19–23 An increase in EIC density with soft CL wear has also been observed in as little as 2 hours.²⁴ However, this immediate response to CL wear has only been examined at the central cornea and the immune response at other regions, including the peripheral and limbal cornea and conjunctiva remains unexplored. EIC migrate from peripheral corneal regions to the central cornea,^25–27 therefore, EIC changes in these regions can provide a more comprehensive view of EIC migratory patterns during CL wear. 

EIC morphology, including cell body size, coupled with changes in the presence and appearance of dendrites, provides insights into EIC antigen capture and migratory capacity in vivo.¹⁸ Studies in mouse models of the dry eye show EIC activation reflected in phenotypic expression and morphology changes at both the central and peripheral cornea within 2 hours of dry eye onset.²⁸ In the human corneal epithelium, an increase in EIC dendrite length, cell area, and number of dendrites has been observed after 1 week of soft CL wear,^21,22 but more immediate changes in EIC morphology in response to CL wear remain unexplored. Given that the initial step in immune activation involves EIC migration, it is anticipated that changes to EIC morphology occur earlier or simultaneously with EIC density change.²⁹ EIC morphology changes at the limbal cornea and conjunctiva have not yet been examined in response to CL wear. 

This study aimed to examine the immediate ocular surface immune response to soft CL wear by observing EIC density and morphology at the central, peripheral, limbal cornea, and bulbar conjunctiva prior to and after 2 hours of CL wear. The relationship between EIC density and morphology during this time was also examined. 

Methods

Study Design

This was a prospective, observational clinical study conducted at the University of New South Wales Sydney (UNSW Sydney). Approval was obtained from the UNSW Human Research Ethics Committee (HC190476). The study followed the tenets of the Declaration of Helsinki and informed consent was obtained from all participants before enrollment. 

Participants

Participants aged between 8 and 42 years and were recruited from UNSW and the local community. The primary inclusion criteria involved being naïve to CL wear, having no active corneal infection or inflammation, including dry eye disease, no previous ocular injuries or surgeries, no corneal irregularities, no current use of eye drops or ocular medication, no systemic or ocular disease likely to affect the ocular surface (e.g. thyroid eye disease, autoimmune disease, and diabetes mellitus), not taking associated medications, and not being pregnant. 

Sample size estimates were based on EIC density data obtained from a meta-analysis conducted on non-CL wearers,¹¹ as well as the published data for 2 hours of CL wear.²⁴ Forty-eight participants would be required for a power of 80% and a 95% of confidence level to detect a change in EIC density of 10 cells/mm² (SAS software version 9.4; SAS, Cary, NC, USA). To account for the possibility of a non-normal distribution in the primary outcome variable, a total of 54 participants were recruited. 

Study Procedures

In Vivo Confocal Microscopy

IVCM was conducted before lens insertion (baseline) and after lens removal (2 hours) on the right eye of each participant using the Heidelberg Retina Tomograph III with Rostock Cornea Module (Heidelberg Engineering GmbH, Heidelberg, Germany). A sterile disposable Perspex cap (Tomocap, Heidelberg, Germany) was used for each participant with a carbomer gel (Viscotears Liquid Gel, Alcon, Australia) as a coupling agent. A non-preserved anesthetic eye drop (oxybuprocaine hydrochloride 0.4%; Chauvin Pharmaceuticals Ltd., London, UK) was instilled into each eye before the examination. 

Images of EIC were captured at the central corneal region, the temporal peripheral cornea (9 o'clock position, identified by the presence of corneal nerves approximately 1–2 mm inside the limbal cornea; Fig. 1), the temporal limbal cornea (confirmed by the observation of palisades of Vogt; see Fig. 1) and the temporal bulbar conjunctiva (approximately 2 mm from the limbal cornea), as described earlier.¹⁰ The corneal sub-basal epithelium was detected at 35 to 70 µm, whereas the conjunctival epithelium was detected at a depth of 5 to 20 µm. Approximately 200 IVCM (400 × 400 µm) images were captured from each location for each participant using the section mode of the IVCM. 

Figure 1.

 

Representative figure depicting the ocular region's localization using in vivo confocal microscopy. The peripheral cornea was defined as the region located approximately 1 to 2 mm inside the limbus, and corneal nerves were prominently observed as landmarks (black arrow). The palisades of Vogt were essential anatomic references utilized to identify the limbal cornea (white arrow), whereas superficial conjunctival cells with bright white nuclei served as distinctive markers for identifying the bulbar conjunctiva (dashed arrow).

Contact Lenses

Immediately after confocal microscopy, soft CLs were inserted and then removed after 2 hours and IVCM was conducted again. The 1-Day ACUVUE MOIST daily disposable soft CL (Etafilcon A, Johnson & Johnson, New Brunswick, NJ, USA; 14.2 mm diameter, 8.5 mm base curve, −0.50 diopters [D]) were worn in both eyes. 

Assessment of EIC Density and Morphology

Five highest quality, best-focused, and less than 20% overlapping IVCM images were manually selected from each region before and after CL wear. EIC were identified as bright/hyper-reflective cells at least 10 µm in size, with a linear/curvilinear cell body, with or without dendrites (both short and long), located at the level of the corneal sub-basal epithelium.¹⁸ Images were analyzed by a single observer (author R.M.) masked to lens wear status (baseline or 2 hours) and participant demographics. EIC were counted manually in the five images and mean values were used for EIC density. The highest grade EIC morphology across the five images was recorded using a published grading system for cell body size, presence of dendrites, presence of long dendrites, and presence of thick dendrites, as described in detail elsewhere.¹⁸ Images without EIC were excluded from the morphology analyses. 

Data Analysis

The normality of EIC density was assessed using the Shapiro-Wilk test (P > 0.05) and by visualization of q-q plots. The EIC density data are continuous and presented as the median along with the interquartile range (cell/mm²). Data related to cell body size are ordinal and expressed as percentages representing small, medium, and large cell bodies. Whereas the data concerning dendrite presence is binary and presented as percentages indicating the presence of dendrites. IBM SPSS (Software version 28; SPSS Inc., Chicago, IL, USA) was used for all data analyses. P values < 0.05 denoted statistical significance. 

Effect of Contact Lens Wear on EIC Density and Morphology

To examine the effect of CL wear on EIC density and morphology, a generalized estimating equation (GEE) model with a fixed effect of CL wear status was conducted for each EIC parameter separately. The linear function of the GEE model was used for EIC density, the ordinal logistic function for cell body size, and the binary logistic function for the presence of dendrites, the presence of long dendrites, and the presence of thick dendrites. As EIC density was not normally distributed (P < 0.001), to adjust for the skewed distribution, EIC density was converted to log units before inclusion in the model. To examine if the change in EIC parameters with CL wear was specific to any ocular surface region, the interaction between CL wear status and the ocular region was examined in the same GEE model. Where the interaction was significant, the effect of CL wear was examined at each region separately using paired sample t-tests (density), Wilcoxon signed rank test (cell body size), and McNemar test (presence of dendrites). In the absence of significant interaction, the effect of CL wear was considered for all regions combined and the results were adjusted for repeat measurements. 

Relationships Between EIC Density and Morphology

Kruskal Wallis and Mann-Whitney U tests were used to compare EIC density between cell body sizes (small, medium, and large) and dendrites (presence and absence). To determine if these differences were specific to CL wear status (baseline or 2 hours) or one ocular region, a GEE model was conducted for EIC density with morphology as a covariate. Interactions of CL wear status and ocular region with differences in EIC density were examined. If the interaction was significant, differences were examined separately for each study visit or each ocular region; otherwise, differences were examined for both study visits or all ocular regions combined and corrected for repeat measurements. 

Results

Fifty-four participants (mean age = 24.8 ± 9.8 years, range = 8–44 years, 44% female participants) completed the study. No adverse effects or corneal staining was observed in any participant with CL wear. EICs were observed at the peripheral and limbal cornea in all participants but were absent at the central cornea in four participants and in almost half (25) of the participants at the conjunctiva. Representative confocal images from each ocular surface region before and after CL wear are provided in Figure 2. Data for EIC density and morphology are presented in Figures 3 and 4 and Table 1. The results of the GEE model are provided in Table 2. 

Table 1.

 

Epithelial Immune Cell (EIC) Morphology: Presence of Dendrites, Long Dendrites, and Thick Dendrites at the Central Cornea, Temporal Peripheral Cornea, Temporal Limbal Cornea, and Temporal Bulbar Conjunctiva Before (Baseline) and After 2 Hours of Soft Contact Lens (CL) Wear

 

Table 2.

 

Summaries of Generalized Estimating Equation (GEE) Model Results for Effects of Contact Lens (CL) Wear on Epithelial Immune Cell (EIC) Density and Morphology

 

Effect of Contact Lens Wear on EIC Density and Morphology

After 2 hours of CL wear, there was a significant increase in EIC density at the peripheral cornea, the limbal cornea, and the conjunctiva, but no change at the central cornea (P = 0.88; see Fig. 3). After 2 hours of CL wear, cell body size significantly increased at the central and peripheral cornea (see Fig. 4), and significantly more participants displayed EICs with long dendrites at these corneal regions (see Table 1). Significantly more participants displayed EICs with thick dendrites at the peripheral (P = 0.04) and limbal cornea (P < 0.001) after CL wear (see Table 1). There was no change in the presence of dendrites after 2 hours of CL wear (P = 0.88). There was no change in conjunctival EIC morphology after CL wear (P ≥ 0.2). 

Figure 2.

 

Representative in vivo confocal microscope images of epithelial immune cells (EICs) at the central cornea, temporal peripheral cornea, temporal limbal cornea, and temporal bulbar conjunctiva. The upper panel represents the cells at baseline and the lower panel represents cells after 2 hours of soft contact lens wear. Image size is 400 × 400 µm² and scale bar indicates 50 µm. (White arrows = corneal nerves, black arrows = EIC, white dashed arrows = EIC with long dendrite, and black dashed arrows = EIC with thick dendrites).

Figure 3.

 

Epithelial immune cell (EIC) density at the central cornea, temporal peripheral cornea, temporal limbal cornea, and temporal bulbar conjunctiva before and after 2 hours of soft contact lens (CL) wear. Plots represent the median (horizontal black line), interquartile range (box), lower and upper extremes Q1/Q3 + 1.5 × interquartile range (whiskers) and outliers lying above Q3 + 1.5 × interquartile range (circles).

Figure 4.

 

Epithelial immune cell (EIC) body size at the central cornea, temporal peripheral cornea, temporal limbal cornea, and temporal bulbar conjunctiva before (baseline) and after 2 hours of soft contact lens (CL) wear. Images devoid of EIC are not presented and were excluded from analyses.

Relationships Between EIC Density and Morphology

There was a significant interaction of the difference in EIC density with the ocular regions (P < 0.001) but not CL wear status. EIC density was found to differ significantly with cell body size at the central (P < 0.001) and limbal corneas (P < 0.001; Fig. 5). At the central cornea, the density of EICs with large cell bodies was higher than EICs with small cell bodies (P < 0.001). Conversely, at the limbal cornea, density of EICs with small cell bodies was higher than that of EICs with large or medium cell bodies (both P < 0.001; see Fig. 5). At the central cornea, the density of EICs with dendrites (P < 0.001) and with long dendrites (P < 0.001) was higher compared to EICs without dendrites (see Fig. 5). At the limbal cornea, the density of EICs with thick dendrites was higher than EICs without thick dendrites (P < 0.001; see Fig. 5). There was no relationship between EIC density and morphology at the peripheral cornea and bulbar conjunctiva. 

Figure 5.

 

Boxplots illustrating the relationships between epithelial immune cell (EIC) density and morphology at the central (A–C) and limbal cornea (D, E). Panels (A–C) focus on the central cornea, examining EIC density with cell body size (A), presence of dendrites (B), and presence of long dendrites (C). Panels (D, E) focus on the limbal cornea examining EIC density with cell body size (D), and the presence of thick dendrites.

Discussion

This study represents the first comprehensive investigation of the immediate response of presumed corneal EIC following soft CL wear in the central, peripheral and limbal cornea, and bulbar conjunctiva. There was an increase in EIC density across all regions, except for the central cornea. After 2 hours of CL wear, changes were observed in EIC cell body size, and the presence and appearance of dendrites, suggesting an increase in antigen-capture capacity throughout the cornea and enhanced migratory capacity of cells in the peripheral and limbal cornea. These findings suggest that soft CL wear triggers an immediate EIC response which may lead to subsequent immune response activation. 

The current study shows for the first time increased EIC density at the peripheral and limbal cornea as well as conjunctiva following short-term CL wear. After 2 hours of CL wear, there was no change in EIC density at the central cornea, which is broadly similar to a study reporting EIC density following short-term (3 hours) CL wear at the central cornea and inferior whorl regions in naïve wearers.³⁰ Conversely, an earlier report showed a twofold increase in EIC density after the same wear duration.²⁴ Differences may be attributed to variations in inclusion criteria. The first two studies used subjects naïve to CL wear, whereas the third study allowed individuals with a history of CL wear to participate, although with a criterion of no CL wear for 6 months prior to enrollment.²⁴ EIC density, however, remains high even after cessation of CL wear for up to 6 months,²¹ perhaps suggesting an adaptive response to CL wear. Density changes in EIC at peripheral regions without corresponding changes at the central cornea might suggest that naïve participants perhaps react slower than those with prior exposure to CL wear and/or a longer duration of wear may have confirmed changes in the central cornea. 

No studies have directly compared the EIC response to soft CL wear at the peripheral ocular regions over similar wear durations. However, increased EIC density at the peripheral cornea, but not the central cornea, has been reported after 1 week of lens wear.^21,22 Animal models show EIC density changes with CL wear occurring first at the peripheral cornea followed by the central cornea.³¹ It is also hypothesized that the increased density of EIC in inflamed eyes is consequential to cells migrating from the draining lymph nodes via limbal vessels to the periphery and then to the center of the cornea.²⁷ Our findings are consistent with these observations, indicating that cells accumulate in peripheral regions first and later reach the central cornea. 

This study presents in vivo EIC morphology changes at the central cornea, peripheral, limbal cornea, and conjunctiva during CL wear. The findings indicate an increased antigen capture capacity at both the central and peripheral corneal regions. Additionally, at the central cornea, the density of EIC with large cell bodies and with dendrites and long dendrites was higher after CL wear. An early IVCM report has observed cells with visible dendrites at the central cornea following CL wear.¹⁹ EIC morphological changes after one week of CL wear at the central and peripheral cornea include significant increases in dendrite length, area, and number of dendrites.²¹ Similar qualitative results have been reported during CL wear in hamsters at the central, peripheral, and limbal cornea and these findings collectively align with our observations.³¹ Although these studies have not explicitly linked morphology to EIC function, our grading system¹⁸ provides meaningful interpretation for these observed morphological changes and suggests that CL wear immediately activates EIC, enhancing their antigen capture capacity in the cornea. 

The findings in this study suggest an augmented EIC migratory capacity at the peripheral and limbal cornea. Additionally, the density of EIC with smaller cell bodies and thick dendrites was higher at the limbal cornea. The corneal limbus has a key role in maintaining the integrity of the corneal epithelium and facilitating wound healing.^32–34 The highest density of EIC at the limbus has been confirmed by other IVCM reports^13,35 and the greater migratory capacity of limbal EIC has also been confirmed in mice.³⁶ The observed distribution is likely that the limbal cornea harbors resident EIC that readily migrate in response to a corneal challenge.^6,25,27 It can be postulated that cells at the limbal cornea are primed to migrate to facilitate the transfer or capture of antigens, whereas at the peripheral cornea, their role is primarily capturing antigens with the potential to migrate if needed to reach the central cornea or limbus.³⁶ This observation of increased migratory capacity and density of EIC at the limbal and peripheral cornea aligns with dendritic cell changes seen in ex vivo murine studies^28,36 and likely reflects the role of dendritic cells in the activation of the immune response.^37–40 It is therefore hypothesized that activation of the early immune response to CL wear in naïve wearers is similarly mediated by dendritic cells, as seen in an animal model of CL wear.⁴¹ 

Consistent with the literature, the lowest number of EICs was observed in the conjunctival epithelium compared to other ocular regions.¹⁰ A higher density of dendritic cells is observed in the conjunctival stroma compared to the epithelium.^42,43 EIC at the limbus generally reside close to the vasculature,⁶ whereas the conjunctival blood vessels are seen in the stroma,⁴² however, further investigation is warranted to determine the precise distribution of dendritic cells in the conjunctiva. 

EIC morphology at the conjunctiva has not been explored previously during CL wear, and, in the current study, no significant changes in EIC morphology were observed, although cell morphology did appear different to the cornea. The lack of difference with CL wear may result from an inherent limitation in our grading system,¹⁸ which was primarily designed for assessing corneal EIC and may not be sufficiently sensitive for evaluating conjunctival EIC. Conceivably, the loose conjunctival tissue compared with the tightly arranged corneal tissue may result in different cell morphology between the structures, or conjunctival cells may belong to a different subtype or possess distinct characteristics, however, this needs further investigation. 

The main strength of this study is the examination of the impact of short-term CL wear in an appropriately powered exploration in a carefully curated cohort, built with an extensive list of exclusion criteria. Participants were completely naïve to CL wear and were free from any ocular disease that may have an underlying inflammatory nature. Additionally, the CL design and refractive power were constant across all participants to exclude other sources of variation. 

A limitation of this study is the reliance on cell morphology for identification of immune cells, which is, however, a widely used approach in vivo,^{13,19,44–46} rather than full cell characterization, which is possible only in vitro. Two recent publications using a combination of flow cytometry and immunohistochemistry in murine models and in human donor corneas, characterized cell type in vitro and used human time-lapse IVCM to explore cell behavior in vivo.^30,47 Cells seen in IVCM are likely to be a combination of T cell and dendritic cell populations,^30,47 where small, mobile immune cells lacking visible dendrites are believed to be T cells, whereas larger cells with evident dendrites are dendritic cells.³⁰ We believe these observations do not impact the findings of the current study. The morphology grading system used in the current study¹⁸ reports size and morphological characteristics of cells at least 10 µm in cell body diameter and separately considers cells lacking discernible dendrites. The type of dendrites is then reported only for those cells where dendrites are visible, which allows analysis of morphologically distinct immune cell populations. A recent animal study investigated the cells involved in subclinical inflammation in the cornea during contact lens wear.⁴⁸ This study revealed that the cells involved in this process were dendritic cells, and some of the cells found at the periphery were T cells.⁴⁸ T cells are considered to be 10 to 20 µm in size,^49,50 and, consequently, the EIC defined in the small cells identified in the current study could be a combination of T cells and dendritic cells, whereas the medium or large cells are representative of dendritic cells. The current study does not classify immune cells as mature or immature^46,51 and instead characterizes “antigen-capture capacity” and “migratory capacity” based on morphology, which we believe to reflect the functional ability in vivo. 

Despite the attempt to control all variables with a careful selection of participants, there are inherent weaknesses within IVCM studies involving repeated analyses. For the peripheral and central cornea, corneal nerves were used as landmarks,⁵² whereas palisades of Vogt helped precise identification of the limbal cornea so that IVCM images from before and after CL wear could be matched as closely as possible. However, it was not possible to use landmarks and precisely locate the imaging region for the conjunctiva. Techniques to enable repeatable analyses of the same location for the conjunctiva would be useful in limiting variability. 

This study investigated the immediate response to CL wear in terms of EIC density, morphology, and distribution. The enhanced antigen-capture and migratory capacity of EIC, coupled with EIC density changes, possibly indicates the initiation of EIC activation or migration into the cornea after soft CL wear that can lead to paraclinical inflammation. These initial alterations in the ocular surface EIC, occurring as a part of the subclinical response to CL wear, are associated with modifications in tear inflammatory mediators.⁵³ Consequently, it is important to investigate the corresponding changes in tear inflammatory markers in immediate CL wear. For future studies, the period of wear should be extended to determine when EICs were able to reach the central cornea or whether EIC morphology characteristics remain the same or alter further with CL wear. Lastly, the implications of these EIC changes should be further studied specifically during the course of corneal infiltrative events. 

Acknowledgments

Funded by Joe and Janet Barr Early Career Cornea and Contact Lens Research Award by the American Academy of Optometry. 

Disclosure: R. Mobeen, None; F. Stapleton, None; C. Chao, None; M.C. Huynh, None; Y.S. Phoebe Wong, None; T. Naduvilath, None; B. Golebiowski, None