Predisposing Factors for Solution-Induced Corneal Staining

Young, Graeme*; Canavan, Kristy; Jones, Susanna; Hunt, Chris§

doi: 10.1097/OPX.0b013e31826ca27e
Original Articles

Purpose: To evaluate predisposing factors in subjects susceptible to solution-induced corneal staining (SICS).

Methods: Four hundred forty-six existing soft contact lens wearers were randomly fitted with one of two study lenses (lotrafilcon B, CIBA Vision, or omafilcon A, CooperVision) and used a hydrogen peroxide care system (Clear Care, CIBAVision) for a 1-month period. Each subject was then challenged with the same lenses presoaked in a multipurpose solution (ReNu MultiPlus, Bausch & Lomb) and assessed after approximately 2 h to be categorized as either “stainers” or “non-stainers” based on whether they showed a classic SICS pattern (diffuse punctate corneal staining in three or more peripheral quadrants).

Results: Overall, 36% (155 of 425) of eligible subjects were classified as stainers and included a higher proportion of omafilcon wearers: 44% (91 of 205) vs. 29% (64 of 220). After using the peroxide system, the stainers were more likely to show poorer wetting (odds ratio [OR] = 1.76, p = 0.02) compared with the non-stainers. Lens wetting showed a significant negative correlation with film deposits (r = −0.63, p < 0.0001, n = 425). Furthermore, on slit lamp examination, the stainers were more likely to show higher levels of upper- and lower-palpebral hyperemia (OR = 2.18, p = 0.02; OR = 1.93, p = 0.04), tarsal roughness (OR = 1.71, p = 0.01), corneal staining (OR = 17.1, p < 0.0001), and conjunctival staining (OR = 1.94, p = 0.004). Although not exposed to preservatives, twice as many eyes in the stainers group had some level of corneal staining: 51% vs. 25%. There were no significant differences between the stainers and non-stainers with respect to age, sex, refraction, or keratometry (p > 0.05).

Conclusions: This study highlighted some notable trends in patients susceptible to SICS. The findings point toward poor tear quality as a predisposing factor for SICS.

*MPhil, PhD, FCOptom, FAAO




Visioncare Research Ltd., Farnham, United Kingdom (GY, SJ, CH), and Vistakon, Division of Johnson & Johnson Vision Care Inc., Jacksonville, Florida (KC).

Article Outline

Contact lens care products containing polyhexamethylene biguanide (PHMB) are used by a high proportion of soft contact lens wearers in the United States, Canada, and the United Kingdom. However, contact lens disinfection solutions that contain PHMB are known to cause punctate corneal staining.1–13 Because the staining is usually asymptomatic and disappears within a couple of hours, some commentators consider this staining to be clinically insignificant.14–16 Hydrogen peroxide care systems are known to result in minimal levels of corneal staining.11,17,18

Solution-induced corneal staining (SICS) can be distinguished from other causes of staining by its characteristic diffuse punctate staining pattern, which, unlike other types of staining, is usually evident in three or more peripheral corneal sectors. Often, the punctate staining follows a circular pattern, but in more severe cases, can involve the whole of the cornea. Generally, the level of corneal staining in SICS is low grade, although more severe cases may be associated with reduced comfort12,13 and an increased incidence of infiltrative events.17 In those subjects showing SICS affecting the entire corneal surface, a three times greater level of infiltrative events has been noted.

The signs of SICS become evident approximately 1 to 2 h after lens insertion9 and usually disappear 4 to 5 h after insertion.10 SICS appears to be related to increased sloughing of corneal epithelial cells.19 It is likely, therefore, that a high proportion of cases of SICS pass unnoticed at routine aftercare visits.

Other longer-term changes such as increased palpebral roughness20 and reduced corneal sensitivity21 have also been noted with PHMB solutions; however, these have not been directly correlated with corneal staining.

The level of staining varies between lens materials, with some silicone hydrogel materials showing high levels of staining.3,5,8 Those giving the highest prevalence of SICS include group 2 hydrogels and plasma-coated silicone hydrogels when used with certain PHMB products. Andrasko and Ryen found that approximately three-quarters of subjects showed SICS with combinations of balafilcon A (PureVision, Bausch & Lomb) + ReNu MultiPlus (RNM; Bausch & Lomb; 73%) and alphafilcon A (Soflens 66, Bausch & Lomb) + RNM (73%).11 With other combinations, the prevalence is much lower; for instance, with etafilcon A (Acuvue 2, Johnson & Johnson Vision Care) + RNM, the prevalence is 1%.

Some patients appear to be more susceptible to SICS than others. When challenged with a given lens-solution combination under controlled conditions, some patients will demonstrate classic SICS, whereas others do not show any staining. The differences in response possibly depend on lens fit, tear film quality, or robustness of corneal epithelium. The purpose of this study was to identify those subjects who are susceptible to SICS, with a view to identifying predisposing factors.

Back to Top | Article Outline


This was an open-label, 1-month, cohort study to identify soft lens wearers susceptible to SICS. Each subject was randomly fitted with one of two study lenses (Table 1): lotrafilcon B (Air Optix, Ciba Vision) or omafilcon A (Proclear, CooperVision), and used a hydrogen peroxide disinfection system (Clear Care, Ciba Vision) for 1 month (Table 2). After a 1-month period of daily wear, the subjects were reexamined and then challenged with the same lenses presoaked in a care system (ReNu MultiPlus Multi-Purpose solution). These lens-solution combinations are known to result in relatively high levels of SICS.11

Four hundred forty-six subjects were recruited at 25 investigational sites in North America (23) and the United Kingdom (2). The study conformed to the tenets of the Declaration of Helsinki; the protocol was reviewed by local institutional review boards; and informed consent was obtained from each subject before commencing the study. All the subjects were existing successful users of daily-wear soft contact lenses. They were required to be at least 18 years of age, and refractive astigmatism, if present, was required to be ≤1.00 D in both eyes.

At the baseline visit, the subjects were examined to ensure good ocular health, and a variety of additional clinical assessments were made in a predetermined sequence (Table 3). Subjects were asked to rate lens comfort on an 11-point scale, where 0 = painful and 10 = unable to feel the lens. Subjects also reported symptoms using a widely used dry eye questionnaire, which asked about the frequency and intensity (on a scale of 1 to 5) of nine dryness-related symptoms.22 Lens surface characteristics were assessed using a slit lamp and included film deposits, white spot deposits, and surface wettability. Film deposits were assessed on a scale of 0 to 4 (where 0 = none and 4 = heavy film), whereas for any white spot deposits, the number of spots were counted. Lens surface wetting was assessed on a scale of 0 to 4, where 0 = very poor, 2 = acceptable (drying time > interblink period), and 4 = excellent, as per normal preocular tear film. Limbal, bulbar, and palpebral hyperemia were assessed on a scale of 0 to 4, where 0 = none and 4 = severe. Tarsal roughness was assessed with white light using the 8-point scale of Løfstrøm et al.,20 where 0 = smooth, 4 = small papillae, and 7 = large papillae. Corneal staining was assessed using blue light and a yellow filter and graded in each of five corneal sectors for severity on a scale of 0 to 4 and for area of staining on a scale of 0 to 10.10 Conjunctival staining was graded on a scale of 0 to 4, where 0 = none and 4 = deep confluent staining.

Eligible subjects were randomly assigned to wear one of the two study lenses using a sequential randomization log. Subjects were instructed in the use of the Clear Care hydrogen peroxide disinfection system and told to use no other care system before the next visit. The lenses were required to show acceptable fit based on acceptable centration, postblink movement, and tightness on push-up; the latter two were assessed on a scale of 0 to 4, where 0 = excessive movement and 4 = no movement.

Where thought to be appropriate by the investigator, subjects were reviewed at a 2-week check-up visit, otherwise all subjects were seen at a 1-month visit. This visit was scheduled 30 days (±5 days) from the dispensing date. This effectively acted as a washout period from lens care preservatives.

The 1-month preassessment visit was undertaken to evaluate the test lenses and ocular physiology. After this, new lenses were inserted that had been presoaked in RNM. These lenses had been soaked in RNM for 18 to 72 h. The lens cases used for presoaking had been conditioned by precycling in RNM for approximately 24 h.

The staining assessment visit was scheduled for 2 to 4 h after insertion. At this visit, subjects were re-questioned about comfort and symptoms, and ocular physiology was rechecked. In particular, corneal staining with fluorescein was evaluated using blue light and a yellow filter (Lees No. 104, Lees, Andover, United Kingdom).

Based on the findings of this challenge, subjects were categorized as either “stainers” or “non-stainers.” Stainers were defined as those subjects susceptible to SICS through showing diffuse punctate staining in three or four peripheral sectors in one or both eyes. It is possible for patients showing “smile-pattern” staining to show punctate staining in three peripheral sectors (inferior, nasal, and temporal) owing to desiccation. However, these cases are usually easily differentiated by their linear pattern of the staining, which corresponds to the lower lid. Subjects showing merely desiccation staining were classified as non-stainers.

Mixed-model analysis was performed with continuous variable data and was used to compare stainers vs. non-stainers at the baseline and challenge visits. The model included stainers group and lens type as fixed effects and site, subject (nested within site), and, where applicable, eye (nested in subject) as random effects. In addition, for the 1-month prechallenge visit, a generalized linear mixed model with a binomial distribution was used to calculate the odds ratio (OR) of being a stainer vs. non-stainer.

For dichotomous variables, the χ2 test was used to compare stainers vs. non-stainers at the baseline visit.

Spearman rank correlations were used to identify relationships between selected variables. For the correlation analysis after 1-month peroxide usage, data from the left and right eyes were averaged.

Back to Top | Article Outline


Of the 446 subjects enrolled, 21 (4%) discontinued before the staining assessment (Fig. 1). The most common reason for discontinuation was lens-related discomfort (six lotrafilcon B, two omafilcon A); three lotrafilcon B wearers discontinued for poor vision. The remaining 10 subjects discontinued for non-clinical reasons (i.e., lost to follow-up, pregnancy, inability to attend visits).

Subjects’ baseline biometric data are summarized in Table 4. The mean spherical refraction was −3.18 D and ranged from −10.50 to + 4.00 D. Mean cylindrical refraction was 0.30 D, and the maximum cylindrical refraction was 1.50 D.

All the subjects were existing soft contact lens wearers, with the majority (82%) being habitual wearers of monthly or 2-weekly replacement lenses. The most commonly worn previous lens type was Acuvue Oasys (23%), followed by Acuvue 2 (11%) and Air Optix (10%).

A high proportion of subjects habitually used multipurpose solutions. The most commonly used habitual care system was Opti-Free RepleniSH (28%), followed by ReNu (14%) and Opti-Free Express (12%). Only a small proportion of subjects (6%) were existing hydrogen peroxide users, and a small number of subjects (5%) were daily disposable lens wearers and therefore used no care system.

As noted earlier, for the data analysis, subjects were categorized from the staining assessment as either a “stainer” or “non-stainer” based on the presence of diffuse punctate staining in three or more peripheral sectors. Sixteen subjects showed this pattern of SICS staining in one eye and little or no staining in the other but were categorized as stainers.

Overall, 36% (155 of 425) of eligible subjects were classified as stainers. A higher proportion of the omafilcon A group were classified as stainers: 44% (91 of 205) compared with 29% (64 of 220) from the lotrafilcon B group (p = 0.0012).

There were no significant differences between the stainer and non-stainers groups with respect to age, sex, sphere refraction, astigmatism, or keratometry (Table 4).

Back to Top | Article Outline

Stainers vs. Non-Stainers after 1-Month Peroxide Usage

After 1-month peroxide usage (i.e., before the RNM challenge), there was no significant difference between stainers and non-stainers in subjects’ assessment of lens-wearing comfort (Table 5) or with any of the symptoms. With both groups, approximately one-half of the subjects noted some dryness.

The stainers were graded as showing significantly poorer lens surface wetting (Table 5). The mean surface wetting scores were 3.3 and 3.5 (on a scale of 0 to 4) for stainers and non-stainers, respectively. A higher proportion of stainers showed less than optimum surface wetting (i.e., grade <4) than the non-stainers: 60% vs. 45%. The OR of 1.76 (p = 0.02) indicates that eyes with surface wetting one grade poorer were nearly twice more likely to be stainers than non-stainers.

A slightly higher proportion of stainers showed some film deposit compared with non-stainers: 45% vs. 37% (OR = 1.49, p = 0.051). There was no significant difference in the level of white spot deposits.

To better understand the implications of the difference in lens wetting, correlation coefficients were calculated for wetting vs. various other key variables at the 1-month visit (Table 6). Lens wetting showed a significant negative correlation with film deposits (r = −0.63, p < 0.0001, n = 425). Statistically significant positive correlations were also noted with comfort, tarsal roughness, and corneal staining, but with low correlation coefficients (r < 0.20) and can therefore be discounted.

There was no significant difference in visual acuity or lens centration. There were small but statistically significant differences in lens movement and tightness on push-up (p = 0.04, p = 0.02, respectively).

The subjects in the stainers group showed significantly higher levels of upper- and lower-palpebral hyperemia (OR = 2.18, p = 0.02; OR = 1.93, p = 0.04, respectively). Tarsal roughness was also graded higher in the staining group (OR = 1.71, p = 0.01); 69% of eyes in the stainers group showed some tarsal roughness compared with 55% of those in the non-stainers group.

A higher proportion of eyes in the stainers group showed some limbal hyperemia (grade >0; 55% vs. 46%), but this was not significant (p = 0.09). The same was true for bulbar hyperemia (p = 0.09). Correlation coefficients were calculated for limbal hyperemia vs. various other key variables at the 1-month visit after Clear Care usage (Table 7). Significant positive correlations were noted with the following: tarsal roughness (r = +0.51, p < 0.0001, n = 425), bulbar hyperemia (r = +0.84, p < 0.0001), and corneal staining severity (r = +0.23, p < 0.0001). Statistically significant positive correlations were also noted with film deposits and “other” slit lamp findings, but with low correlation coefficients (r < 0.20) and can therefore be discounted.

Although, at this point, the subjects had not been exposed to PHMB, twice as many eyes in the stainers group showed some level of corneal staining as those in the non-stainers group: 51% vs. 25% (OR = 17.1, p < 0.0001, Table 5). Most of the corneal staining was low grade, and none of the subjects showed staining characteristic of SICS. The stainers also showed a significantly higher level of conjunctival staining (OR = 1.04, p = 0.007); 51% of eyes in the stainers group showed some conjunctival staining compared with 37% of those in the non-stainers group.

Interestingly, the stainers also showed a higher level of other slit lamp findings (p = 0.004); 12% (36 of 310) of eyes in the stainers group showed other slit lamp findings compared with 2% (12/540) of those in the non-stainers group. These slit lamp findings included pingueculae (23), meibomian gland dysfunction (6), corneal vascularization (5), corneal scars (4), and concretions (3).

Back to Top | Article Outline

Stainers vs. Non-Stainers After 2 h Challenge

As expected, the stainers showed significantly higher levels of corneal staining than the non-stainers after the RNM challenge (Table 6). Similarly, a larger proportion of eyes in the stainers group showed some level of corneal staining (Grade >0) than in the non-stainers group: 99% vs. 45% (308 of 310 vs. 245 of 540, p < 0.0001). Sixteen subjects showed SICS staining in one eye and little or no staining in the other. This was also the case with most of the other slit lamp findings.

Back to Top | Article Outline


The proportion of stainers noted at the RNM challenge visit was higher with the RNM-Proclear combination than the RNM-Air Optix combination (44% vs. 29%). This is consistent with the findings by Andrasko and Ryen, who also noted a higher prevalence of SICS with the RNM-Proclear combination.11 Although Andrasko and Ryen undertook a similar 2 h challenge test, it is difficult to make a direct comparison with their findings because they did not make a similar stainer/non-stainer judgment, but instead, reported area of staining. However, based on the proportion of subjects with an area of staining >20%, their prevalence of SICS was higher for Proclear than Air Optix (87% and 52%, respectively). It is possible, therefore, that some of the investigators in this study were less effective in identifying corneal staining. However, if this was the case, it possibly means that the study identified the more severe easily recognizable cases and, perhaps, failed to identify borderline stainers.

The observation of SICS in one eye only, in a small proportion of subjects, is one that has been noted in a previous study17: 16% compared with 10% in the present study. A possible explanation for this asymmetrical staining is that, in these instances, one lens contained more residual solution than its fellow lens when inserted into the eye. Another factor may be differences in the amount of lens manipulation; Peterson et al.23 noted a reduction in SICS with balafilcon A lenses after the introduction of a rub-and-rinse step.

An important finding of this study is that patients who are susceptible to SICS tend to show poorer lens wetting, suggesting that one predisposing factor is poor tear quality. It seems possible that, in these eyes, the disinfectant is less dilute and, consequently, the cornea is exposed to a higher dosage of disinfectant. Carnt et al.14 commented that “patients who experienced SICS had poorer lens surface characteristics at dispensing and at the time of the SICS event” but gave no further explanation of what was meant by “lens surface characteristics.” They suggested that this might indicate a relationship between SICS and deposits; however, the short wearing time in the present study would not support this.

The fact that patients susceptible to SICS showed increased corneal staining even before exposure to disinfectant solution may be a side effect of poor tear quality, which, in turn, leads to corneal desiccation staining. Alternatively, this increased staining might be an indication of less robust corneal epithelium, which is more susceptible to corneal staining from a variety of causes. In other words, as well as being more susceptible to SICS, this more fragile corneal epithelium might also be more susceptible to mild mechanical disruption and desiccation.

Despite the apparent effect of wetting and therefore tear film, it was surprising that there were no significant differences in age or sex between subjects susceptible and those not susceptible to SICS.

The additional findings of raised hyperemia and palpebral roughness further add to the impression that those eyes predisposed to SICS are already compromised due to poor tear quality or other factors. Poorer tear film quality, for instance, may result in greater friction between the lid and the lens front surface. This is further emphasized by the higher prevalence of other slit lamp findings, which encompass a range of problems also indicative of ocular compromise and/or poor tears (e.g., pingueculae).

The increased deposition seen in the stainers group would be an expected consequence of poor wetting, and this was confirmed by the correlation between wetting and deposits. Although few studies have successfully correlated deposits with tear film characteristics, it is generally accepted that surface drying is a factor in the spoilation of soft lenses. Zhao et al.24 recently noted an association between SICS and higher levels of protein and cholesterol extracted from worn lenses. Deposits have also been correlated with incomplete blinking.25

A possible criticism of this study is that the significant differences between the two groups were relatively small. However, the aim of the study was to identify trends that might throw light on why some patients are more susceptible to SICS than others. The fact that these differences, albeit small, represent a consistent pattern of ocular compromise suggests a plausible explanation. The use of multiple sites inevitably results in some loss of consistency, but arguably, this makes the findings all the more compelling.

A limitation of this study is that only two combinations of lens and solution were tested. It is likely that other combinations would have resulted in different incidences of SICS, and indeed, the two combinations in this study showed different incidences. However, if the underlying assumption, that some patients are inherently more susceptible to SICS, is correct, it should be possible to apply the trends noted in this study to other lens-solution combinations that generate SICS.

Back to Top | Article Outline


This study has highlighted some notable differences between patients susceptible and not susceptible to SICS. However, these are not clear enough to be able to make reliable predictions as to which patients are susceptible to SICS. The findings point toward marginal dry eye problems as a predisposing factor.

Graeme Young

Visioncare Research Ltd.

Craven House, West Street

Farnham, Surrey GU9 7EN

United Kingdom


Back to Top | Article Outline


We thank the following investigators (in reverse alphabetical order): Richard Wilson, Larry Wan, Joseph Udvari, Dora Sudarsky, Stephen Selden, T. Powers Griffin Jr., Wilson Movic, Stephen Montaquila, Larry Menning, Shane Kannarr, Olivia Hunt, Michael Greenberg, James Fujimoto, James Dugue, Jeanne Derber, Frank D’Apolito, Michael Cymbor, Mark Curtis, Christine Cook, Adam Coffee, David Camp, James Brobst, William Bogus, Peter Benvenuto. The work was supported by Johnson & Johnson Vision Care Inc.

Received November 21, 2011; accepted July 13, 2012.

Back to Top | Article Outline


1. Begley CG, Edrington TB, Chalmers RL. Effect of lens care systems on corneal fluorescein staining and subjective comfort in hydrogel lens wearers. Int Cont Lens Clinic 1994; 21: 7–12.
2. Jones L, Jones D, Houlford M. Clinical comparison of three polyhexanide-preserved multi-purpose contact lens solutions. Cont Lens Anterior Eye 1997; 20: 23–30.
3. Epstein AB. SPK with daily wear of silicone hydrogel lenses and MPS. Contact Lens Spectrum 2002; 17 (11): 30.
4. Fonn D. Observations of corneal staining with MPS and silicone hydrogel lenses. Contact Lens Spectrum 2002; 17 (11): 32.
5. 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.
6. Pritchard N, Young G, Coleman S, Hunt C. Subjective and objective measures of corneal staining related to multipurpose care systems. Cont Lens Anterior Eye 2003; 26: 3–9.
7. Lebow KA, Schachet JL. Evaluation of corneal staining and patient preference with use of three multi-purpose solutions and two brands of soft contact lenses. Eye Contact Lens 2003; 29: 213–20.
8. Amos C. A clinical comparison of two soft lens care systems used with silicone hydrogel contact lenses. Optician 2004; 227: 16–20.
9. Bandamwar KL, Garrett Q, Cheung D, Huang J, Lee L, Ng C, Papas EB. Onset time course of solution induced corneal staining. Cont Lens Anterior Eye 2010; 33: 199–201.
10. Garofalo RJ, Dassanayake N, Carey C, Stein J, Stone R, David R. Corneal staining and subjective symptoms with multipurpose solutions as a function of time. Eye Contact Lens 2005; 31: 166–74.
11. Andrasko G, Ryen K. Corneal staining and comfort observed with traditional and silicone hydrogel lenses and multipurpose solution combinations. Optometry 2008; 79: 444–54.
12. Carnt N, Willcox M, Evans V, Naduvilath T, Tilia D, Papas E, Sweeney D, Holden B. Corneal staining: the IER matrix study. Contact Lens Spectrum 2007; 22 (9): 38–43.
13. Diec J, Evans VE, Tilia D, Naduvilath T, Holden BA, Lazon de la Jara P. Comparison of ocular comfort, vision, and SICS during silicone hydrogel contact lens daily wear. Eye Contact Lens. 2012; 38: 2–6.
14. Carnt N, Evans V, Holden B, Naduvilath T, Tilia D, Papas EB, Willcox MD. IER matrix update: adding another silicone hydrogel. Contact Lens Spectrum 2008; 23 (3): 28–35.
15. Levy B. Superficial corneal “staining”-clinical observation and risk assessment. Eye Contact Lens 2007; 33: 165–6.
16. Levy B, Osborn G. Clinical risks: myths and truths. Contact Lens Spectrum 2008; 23 (1): 42–6.
17. 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.
18. Keir N, Woods CA, Dumbleton K, Jones L. Clinical performance of different care systems with silicone hydrogel contact lenses. Cont Lens Anterior Eye 2010; 33: 189–95.
19. Peterson RC, Gorbet M, Woods CA, Fonn D. The transient nature of solution induced corneal staining. Optom Vis Sci 2009; 86:E-Abstract 090816.
20. Løfstrøm T, Anderson JS, Kruse A. Tarsal abnormalities: a new grading system. CLAO J 1998; 24: 210–15.
21. Epstein AB. Contact lens care products effect on corneal sensitivity and patient comfort. Eye Contact Lens 2006; 32: 128–32.
22. Nichols JJ, Mitchell GL, Nichols KK, Chalmers R, Begley C. The performance of the contact lens dry eye questionnaire as a screening survey for contact lens-related dry eye. Cornea 2002; 21: 469–75.
23. Peterson RC, Fonn D, Woods CA, Jones L. Impact of a rub and rinse on solution-induced corneal staining. Optom Vis Sci 2010; 87: 1030–6.
24. Zhao Z, Naduvilath T, Flanagan JL, Carnt NA, Wei X, Diec J, Evans V, Willcox MD. Contact lens deposits, adverse responses, and clinical ocular surface parameters. Optom Vis Sci 2010; 87: 669–74.
25. Collins MJ, Stahmer D, Pearson G. Clinical findings associated with incomplete blinking in soft lens wearers. Clin Exp Optom 1989; 72: 55–60.

soft contact lens; corneal staining; SICS; wetting; PHMB; hydrogen peroxide; multipurpose solution

© 2012 American Academy of Optometry