Zhao, Zhenjun*; Naduvilath, Thomas*; Flanagan, Judith L.*; Carnt, Nicole A.†; Wei, Xiaojia‡; Diec, Jennie†; Evans, Vicki*; Willcox, Mark D. P.*
Contact lens deposition begins within seconds to minutes of lens insertion into the eye.1,2 This deposition forms a pellicle or film that comprises proteins,1,3 carbohydrates, mucins, lipids,4 from tears as well as from bacteria, and cognate microbial toxins and polysaccharides.5 The degree of deposition from the tear film can depend on age of the lens material, water content of lenses, and lens care solution.6 Resulting alterations in the physical and/or chemical properties of the lens can be associated with discomfort because of decreased wettability7 and reduced visual acuity.8 Tear film components or bacteria adsorbed to lenses may compromise ocular health as, for example, when the components of the pellicle induce immunological/inflammatory responses, which can present as or contribute to contact lens-induced red eye (CLARE)9,10 or papillary conjunctivitis (CLPC).11 Lens deposition may alter tear composition and physiology,12 which may have a further bearing on susceptibility to bacterial infection and perhaps a negative cycle of increased deposition and discomfort.
Silicone hydrogels are reported to exhibit decreased protein accumulation but increased lipid spoliation.13,14 Previously, we have reported the detection of proteins and cholesterol extracted from worn silicone hydrogel lenses and demonstrated the contribution of lens polymer and lens care solution to lipid and protein accumulation.6 In the present report, we extend these investigations to examine correlations between protein and lipid fouling of worn contact lenses, with adverse event rates, clinical variables, and subjective ratings collected at scheduled clinic visits.
MATERIALS AND METHODS
Participants, Contact Lenses, and Lens Care Solutions
All participants were recruited from people who took part in clinical studies conducted within the Institute for Eye Research, Sydney, Australia. Before enrolment in the studies, all participants signed an informed consent after explanation of the nature and possible consequences of the study. Participants were excluded if they had any active ocular infection or inflammation. All experimental protocols were reviewed and approved by the Institutional Human Ethics Review Committee and complied with the Declaration of Helsinki for Experimentation on Humans, 1975 and revised in 1983.
The data analyzed were from studies that were conducted as controlled, prospective, non-randomized, and non-concurrent clinical studies. Wear schedules were daily wear, with lenses being replaced on a monthly or 2 weekly basis. Detailed information about the contact lenses (senofilcon A, balafilcon A, galyfilcon A, lotrafilcon B), lens care solutions [a total of four individual solutions; containing either hydrogen peroxide as disinfecting agent and Pluronic 17R4 as surfactant (ClearCare, CIBA Vision, Atlanta, GA) or Polyhexanide and Pluronic F127 (Aquify, CIBA Vision) or Polyquad/Aldox and Tectronic 1304 (either Opti-Free Express or Opti-Free RepleniSH, Alcon, Fort Worth, TA)], lens wear schedule, lens collection procedure, and amount of protein and lipids deposited on lenses that have been used with different multipurpose contact lens solutions have been reported previously.6,15 In brief, all studies were conducted under three protocols over a period of several years, yet all participants were given the same lens care and wear instructions. The only differences in the protocols were in the replacement periods for lenses. Subjects wore each lens material in combination with each multipurpose disinfecting solution, and there were approximately 40 participants in each lens/solution combination. All subjects attended the clinic at baseline and then after 2 weeks, 1 month, and 3 months of lens wear. However, a subset of these subjects was randomly chosen to have their contact lenses collected and analyzed for protein and lipid deposition.6 Lenses were collected for protein and lipid determination at the 1 month and 3-month visits and each subject contributed one lens. A total of 143 lotrafilcon B lenses, 152 galyfilcon A lenses, 120 senofilcon A lenses, and 168 balafilcon A lenses were collected. Because of the nature of the clinical studies, there were some subjects who participated in more than one lens/solution combination. The statistical modeling (discussed below) took this into account.
Clinical Assessment During Lens Wear
Participants wore lenses in both eyes. Clinical variables were assessed as reported previously.16 Lens surface characteristics including lens wettability, the number of mucin balls, lens fitting performance (primary gaze movement and lens tightness), limbal, bulbar and palpebral redness, subjective ratings of comfort and vision, corneal and conjunctival staining were measured after lenses were removed.16 Lens tightness was measured on a scale of 1 to 100, where 0 corresponded to no tightness (the lens is held in place only by the lid) and 100 corresponded to the lens bound to the cornea. Primary gaze movement (mm) was assessed during normal blinking by observing the lower lens edge. Lens wettability was graded on a scale of 0 to 5, where 0 = totally hydrophobic (non-wetting), 1 = non-wetting patches immediately after blinking, 2 = an appearance equivalent to a hydroxyethylmethylacrylate (HEMA) lens surface, 3 = more wettable than a HEMA lens surface, 4 = an appearance approaching that of a healthy cornea, and 5 = an appearance equivalent to that of a healthy cornea. Grading units were decimalized to improve sensitivity. The number of mucin balls were counted using a biomicroscope under direct white illumination (magnification, ×16). Limbal, bulbar and palpebral redness, and corneal and conjunctival staining were measured using the grading scales from 0 to 4, where 0 = none, 1 = very slight, 2 = slight, 3 = moderate, and 4 = severe. Corneal and conjunctival staining were assessed after the instillation of sodium fluorescein. Visual acuity was measured using wall-mounted Bailey-Lovie logMAR charts. The frequency and types of adverse events and solution-induced corneal staining (SICS) seen with each lens/solution combination have been defined15,17 and reported for these particular lens/solution combinations previously.15 The subjective variables were collected at routine study visits. Participants were asked to rate the comfort of the lenses on insertion, during wear and at the end of the day on scales of 1 to 10, where 1 = very uncomfortable and 10 = extremely comfortable.
Laboratory Analysis of Worn Lens Deposits
A total of 583 worn lenses were collected (at scheduled study visits after 1 month and 3 months of daily wear) and analyzed for deposits. The methods of deposit analysis and the results of protein and cholesterol deposits on worn lenses have been reported previously.6 Briefly, for lipid analysis, lenses were extracted in chloroform:methanol (1:1 v/v) overnight with gentle shaking at ambient temperature, and the non-polar lipids in the extracts separated by thin layer chromatography using hexane followed by benzene followed by hexane:ether:acetic acid (60:40:1 v/v). After drying, the plate was sprayed with 25% sulfuric acid and color developed by heating. For protein analysis, lenses were extracted with 4 M urea, 0.01% (w/v) sodium dodecylsulfate, and 1 mM dithiothreitol (pH, 7.4 at 95°C for 3 h), and protein concentration measured using a bicinchoninic acid or NanoOrange staining assay.
Statistical Analysis of Data
Protein and lipid levels extracted from lenses measured on a continuous scale were the outcome variables of the statistical model. These variables were subject to transformation before data analysis because of the heterogeneity of the variance between the lens-solution groups. The transformation aimed to reduce the differences in variance between groups. Based on the Box-Cox transformation method, the square root transformation was found to be appropriate. As only a subset of lens wearers had the amount of protein and lipid from lenses analyzed, the clinical data of only that subset was used in the statistical analysis.
For analysis of associations between amount of protein or cholesterol extracted from lenses and occurrence of adverse responses or SICS seen during wear, a case/control analysis was performed using linear mixed model. The data for eyes of participants that had an adverse event or SICS at any time were compared to data from eyes of participants that never had the event during the course of the studies. Cases were matched to controls for lens and solution use.
Clinical variables and subjective ratings collected at scheduled clinic visits were analyzed as covariates for associations with the protein and lipid data of their corresponding worn lenses. Univariate associations between variables were determined using Pearson correlation coefficient and slopes for the two separate outcomes, namely proteins and lipids. A correlation coefficient of ≤0.30 was considered to show a small correlation, 0.3 to 0.5 a moderate correlation, and >0.5 a strong correlation. Covariates that were significant at p < 0.2 were considered for multivariate testing. The method of model building comprised initially of backward stepwise removal starting from the most non-significant factor until all variables in the model were significant. This was followed by entering back each excluded covariate to determine any improved value to the model. The final model retained covariates that were significant or confounded other existing factors. Data were analyzed using the linear mixed model, which accounted for correlations arising from repeat observations within participants. The Random intercept was used to define that the intercepts created by these. Unique participant number was random and the covariance within them was defined using a simple covariance structure (scaled identity or variance components). Statistical significance was set at 5%. Data were analyzed using SPSS version 17.
Correlations between Amount of Protein or Cholesterol Extracted from Lenses and Rate of Adverse Events or SICS
Adverse events were analyzed as a single variable [microbial keratitis (MK), contact lens induced acute red eye (CLARE), contact lens-induced peripheral corneal ulcer (CLPU), infiltrative keratitis (IK), superior epithelial arcuate lesion (SEAL), contact lens induced papillary conjunctivitis (CLPC), corneal erosions (CE), asymptomatic IK (AIK), and asymptomatic infiltrates (AI)], or split into the following groups: total corneal infiltrative events (MK, CLARE, CLPU, IK, AI, AIK), significant corneal infiltrative events (MK, CLARE, CLPU, IK), asymptomatic corneal infiltrative events (AI, AIK), and mechanical adverse events (SEAL, CLPC, CE).15 Association of SICS and amount of protein or cholesterol extracted from lenses was analyzed separately. Thirty-seven participants who had their contact lenses measured for protein and 35 who had their lenses measured for lipid had any form of adverse event (MK, CLARE, CLPU, IK, SEAL, CLPC, CE, AIK, or AI). Ten participants who had their contact lenses measured for protein and 13 who had their lenses measured for lipid had any form of corneal infiltrative event (MK, CLARE, CLPU, IK, AI, AIK). Six participants (protein) and seven participants (lipid) had significant corneal infiltrative events. Four participants (protein) and six participants (lipid) had asymptomatic corneal infiltrative events. Seven participants (protein) and seven participants (lipid) had mechanical adverse events. Thirty-seven participants (protein) and 45 participants (lipid) had SICS.
There were significant associations between the amount of protein (p = 0.008) and amount of cholesterol (p = 0.01) extracted from lenses and the level of SICS (Fig. 1), such that people with SICS had higher amounts of protein and lipids extracted from their lenses. However, there was no significant association between amount of protein or cholesterol extracted from lenses and total adverse events (p = 0.52 and 0.07, respectively), total corneal infiltrative events (p = 0.51 and 0.89, respectively), significant corneal infiltrative events (p = 0.78 and 0.9, respectively), asymptomatic corneal infiltrative events (p = 0.49 and 0.95, respectively), or mechanical adverse events (p = 0.2 and 0.07, respectively).
Correlation of Total Protein Deposit and the Clinical Parameters
Univariate correlation analysis of the total protein extracted from worn contact lenses and each of the clinical parameters did not show any moderate or strong correlations (r >0.3, Table 1). But weak correlations (r ≤0.23) existed between the amount of protein extracted and overall corneal staining, overall conjunctiva staining, front surface wetting, and lens fit tightness (Table 1). Overall corneal staining and front surface wetting were positively correlated with the amount of protein extracted from the lenses, whereas overall conjunctival staining and lens fit tightness were negatively correlated. In the multivariate analysis, these clinical parameters accounted for 48% (Table 1) of lens protein deposit, indicating other factors contributed 52% of the deposition (Fig. 2).
In the initial univariate correlation analysis, those factors that were associated with amount of protein extracted from lenses that were not kept within the multivariate analysis included comfort on insertion (R = −0.13, p = 0.033); contact lens centration in the vertical meridian (R = −0.18, p = 0.004), primary gaze movement (r = 0.16, p = 0.01), haziness (r = 0.13, p = 0.03), and level of overall acceptance of contact lens on eye (r = 0.13, p = 0.03). Clinical parameters that failed to exhibit any correlation with total protein deposition included grade of blepharitis, overall meibomian gland blockage, bulbar, limbal or palpebral conjunctival redness, corneal vascularization, conjunctival indentation, conjunctival tarsal abnormalities, grade of number of mucin balls, contact lens centration in the horizontal meridian, primary gaze lag, and comfort during or at end of the day.
Correlation of Cholesterol Deposit and the Clinical Parameters
For cholesterol spoliation, multivariate correlation analysis was performed on different worn contact lenses from the same target population. The amount of cholesterol extracted from lenses and clinical parameters exhibited no strong correlation (r >0.3, Table 2). Two weak positive correlations (r ≤0.19) existed, however, between the amount of cholesterol extracted from lenses and overall cornea vascularization or grade of front surface deposition on lenses (Table 2). However, in the multivariate analysis, these clinical parameters accounted for only 6% of lens amount of cholesterol extracted from lenses.
Univariate correlation analysis variables that were not kept in the multivariate analysis, which correlated with the amount of cholesterol extracted from lenses, included comfort on insertion (R = −0.16, p = 0.008), comfort overall (R = −0.13, p = 0.029), grade of overall conjunctival indentation (R = −0.19, p = 0.013), visual acuity (R = −0.14, p = 0.015), overall grade of corneal staining (r = 0.14, p = 0.018), and number of mucin balls (r = 0.13, p = 0.03). Those factors that did not correlate significantly in the univariate analysis included centration of the lens on eye, bulbar, limbal or palbebral conjunctival redness, grade of blepharitits or meibomium gland blockage, front surface wetting of the lens, haziness of the lens, conjunctival staining, or movement and tightness of the lens on eye.
This study was designed to assess whether protein or cholesterol deposition on silicone hydrogel contact lenses was associated with adverse events seen during contact lens wear, clinical signs of contact lens wear, and comfort of lenses during wear. Previous research suggests that certain adverse events, for example CLPC,11 or discomfort during wear,7 are correlated with deposition, but these previous data were collected for HEMA-based hydrogel lenses and not silicone hydrogel lenses. Indeed, the published findings that silicone hydrogel lenses tend to deposit less protein than HEMA-based hydrogel lenses, but more lipid,13,14,18,19 suggested there may be different clinical effects of the protein or lipid adsorption onto silicone hydrogels. The methodology used in this article may result in confounding factors within the data analysis. It is difficult to disentangle the effect of the material per se and the protein or lipid extracted from that material. It is the material properties, such as surface hydrophobicity/hydrophilicity, that determine the amount of protein/lipid that can be extracted after wear. However, the use of the case/control analysis that was possible when analyzing for effects of protein or lipid on adverse events or SICS allowed us to control for the lens and solution type—the controls were matched to cases for these in the analysis.
An association between the level of SICS and the level of protein or cholesterol that could be extracted from contact lenses after wear was observed. Previously, SICS has only been associated with the type of solution used to disinfect and clean lenses.15,18,20 Staining can occur rapidly (2 h after lens insertion20), within this same time frame in vitro studies have demonstrated both proteins and lipids can deposit onto contact lenses.1,13,14 However, on balance, the clear effect of solution on the rates of SICS20,21 is more likely to be the overall driver of the SICS response, and the association with protein and lipid on lenses is probably correlative rather than causative. SICS has been associated with corneal inflammation21 and decreased comfort,20 and the failure to find a correlation between protein or cholesterol extracted from lenses and corneal inflammation or comfort underscores that these may not be causative. Others have shown an association between lens deposits on HEMA-based hydrogel lenses and comfort.22
This study failed to demonstrate any correlation between the amount of protein or cholesterol extracted from lenses and adverse responses associated with lens wear. Nor was there any correlation between amount of protein or cholesterol and the grade of limbal, bulbar, or palpebral redness (signs of subacute inflammation). The only association with inflammation detected was a weak and probably not physiologically relevant correlation between amount of cholesterol extracted from lenses and the grade corneal vascularization. However, previous research has associated specific proteins, for example albumin,11 or secretory IgA23 with CLPC, so total amount of total protein might not be the appropriate measure to determine such associations. Between 4 and 28 proteins have reported identified in protein extracts from silicone hydrogel lenses23,24 and it is possible that specific proteins of these are associated with, or causative of, adverse events.
In this study, there were weak correlations between amount of protein that could be extracted from lenses after wear and corneal staining and front surface wetting during lens wear (positively correlated), or conjunctival staining and lens tightness during wear (negatively correlated), independent of contribution of individual lens/solution interactions. The weaker correlations between cholesterol that could be extracted after wear, and clinical variables, would suggest that cholesterol deposition does not contribute greatly to clinical or symptomatic changes during lens wear. It should be borne in mind that the correlations reported in this study are between sets of discrete (clinical variables) and continuous (protein/cholesterol levels) data, making strong correlations very unlikely.
In conclusion, this study has highlighted the fact that there may be no direct physiological consequence of cholesterol depositing on lenses during wear as any correlations were very weak. There may be at least a potential for protein deposition on silicone hydrogel lenses to be associated with changes to corneal and conjunctival epithelium (staining), comfort of lenses during wear, and front surface wetting of lenses. However, the correlations were generally small and may still not indicate any causative relevant physiological response. Further work is warranted to elucidate whether specific proteins from the tears contribute to certain correlated findings, and hence whether manufacturers of lenses should consider deposition of proteins in general, or particular proteins and cholesterol, when designing new lenses/lens surfaces that can improve contact lens performance.
We thank our clinical study team for organizing and carrying out the clinical trial.
Mark D. P. Willcox
Institute for Eye Research
University of New South Wales
Sydney, NSW 2052
1. Emch AJ, Nichols JJ. Proteins identified from care solution extractions of silicone hydrogels. Optom Vis Sci 2009;82:123–31.
2. Luensmann D, Jones L. Albumin adsorption to contact lens materials: a review. Cont Lens Anterior Eye 2008;31:179–87.
3. Minarik L, Rapp J. Protein deposits on individual hydrophilic contact lenses: effects of water and ionicity. CLAO J 1989;15:185–8.
4. Bontempo AR, Rapp J. Lipid deposits on hydrophilic and rigid gas permeable contact lenses. CLAO J 1994;20:242–5.
5. Bowers RW, Tighe BJ. Studies of the ocular compatibility of hydrogels. A review of the clinical manifestations of spoilation. Biomaterials 1987;8:83–8.
6. Zhao Z, Carnt NA, Aliwarga Y, Wei X, Naduvilath T, Garrett Q, Korth J, Willcox MD. Care regimen and lens material influence on silicone hydrogel contact lens deposition. Optom Vis Sci 2009;86:251–9.
7. Nilsson SE, Andersson L. Contact lens wear in dry environments. Acta Ophthalmol (Copenh) 1986;64:221–5.
8. Gellatly KW, Brennan NA, Efron N. Visual decrement with deposit accumulation of HEMA contact lenses. Am J Optom Physiol Opt 1988;65:937–41.
9. Holden BA, La Hood D, Grant T, Newton-Howes J, Baleriola-Lucas C, Willcox MD, Sweeney DF. Gram-negative bacteria can induce contact lens related acute red eye (CLARE) responses. CLAO J 1996;22:47–52.
10. Kotow M, Holden BA, Grant T. The value of regular replacement of low water content contact lenses for extended wear. J Am Optom Assoc 1987;58:461–4.
11. Tan ME, Demirci G, Pearce D, Jalbert I, Sankaridurg P, Willcox MD. Contact lens-induced papillary conjunctivitis is associated with increased albumin deposits on extended wear hydrogel lenses. Adv Exp Med Biol 2002;506:951–5.
12. Thai LC, Tomlinson A, Doane MG. Effect of contact lens materials on tear physiology. Optom Vis Sci 2004;81:194–204.
13. Carney FP, Nash WL, Sentell KB. The adsorption of major tear film lipids in vitro to various silicone hydrogels over time. Invest Ophthalmol Vis Sci 2008;49:120–4.
14. Chow LM, Subbaraman LN, Sheardown H, Jones L. Kinetics of in vitro lactoferrin deposition on silicone hydrogel and FDA group II and group IV hydrogel contact lens materials. J Biomater Sci Polym Ed 2009;20:71–82.
15. Carnt NA, Evans VE, Naduvilath TJ, Willcox MD, Papas EB, Frick KD, Holden BA. Contact lens-related adverse events and the silicone hydrogel lenses and daily wear care system used. Arch Ophthalmol 2009;127:1616–23.
16. Stern J, Wong R, Naduvilath TJ, Stretton S, Holden BA, Sweeney DF. Comparison of the performance of 6- or 30-night extended wear schedules with silicone hydrogel lenses over 3 years. Optom Vis Sci 2004;81:398–406.
17. Sweeney DF, Jalbert I, Covey M, Sankaridurg PR, Vajdic C, Holden BA, Sharma S, Ramachandran L, Willcox MD, Rao GN. Clinical characterization of corneal infiltrative events observed with soft contact lens wear. Cornea 2003;22:435–42.
18. Jones L, MacDougall N, Sorbara LG. Asymptomatic corneal staining associated with the use of balafilcon silicone-hydrogel contact lenses disinfected with a polyaminopropyl biguanide-preserved care regimen. Optom Vis Sci 2002;79:753–61.
19. Lorentz H, Jones L. Lipid deposition on hydrogel contact lenses: how history can help us today. Optom Vis Sci 2007;84:286–95.
20. Andrasko G, Ryen K. Corneal staining and comfort observed with traditional and silicone hydrogel lenses and multipurpose solution combinations. Optometry 2008;79:444–54.
21. Carnt N, Jalbert I, Stretton S, Naduvilath T, Papas E. Solution toxicity in soft contact lens daily wear is associated with corneal inflammation. Optom Vis Sci 2007;84:309–15.
22. Jones L, Franklin V, Evans K, Sariri R, Tighe B. Spoilation and clinical performance of monthly vs. three monthly Group II disposable contact lenses. Optom Vis Sci 1996;73:16–21.
23. Zhao Z, Wei X, Aliwarga Y, Carnt NA, Garrett Q, Willcox MD. Proteomic analysis of protein deposits on worn daily wear silicone hydrogel contact lenses. Mol Vis 2008;14:2016–24.
24. Green-Church KB, Nichols JJ. Mass spectrometry-based proteomic analyses of contact lens deposition. Mol Vis 2008;14:291–7.