Aspects of Corneal Fluorescein Staining

Abstract

PURPOSE Evaluating the cornea for epithelial fluorescein staining is a key element of the ocular examination of contact lens wearers and people with dry eye disease. It has long been viewed as a method of visualizing a break in the protective epithelial layer, the integrity of which is regarded as vital for protecting the eye and maintaining good vision. There has been very little reported on the typical staining presentation in dry eye. Understanding more about the distribution of epithelial staining in dry eye disease would be valuable to guide evaluation of a treatment’s physiological efficacy. This thesis aimed to determine whether the corneal staining of subjects with symptomatic dry eye presents in a specific distribution pattern Since 2002, the epithelial staining phenomenon of solution induced corneal staining (SICS) has been investigated. The cause of this staining has been suggested to be due to molecular adhesion rather than physiological damage, but the current evidence is equivocal. More investigation of this phenomenon is warranted to understand the process and the clinical significance of SICS. This thesis aimed to investigate the type, severity and pattern of staining that occurs in SICS, and assess the impact on epithelial cells using in-vivo confocal imaging.

METHODS Chapter 2 described the CORE corneal staining scale., which uniquely reports the type and extent of the corneal staining on a scale of 0-100. This was the staining scale used to record the level of solution induced corneal staining in all the clinical trials featured in Chapters 5 and 6. Chapter 3 reported an experiment which was conducted to assess the agreement among fifteen observers who used this scale, in two grading sessions, to grade the corneal staining illustrated in 22 photographic images. Inter- and intra-observer agreement results were calculated. Chapter 4 presented a meta-analysis of the corneal staining observed in 368 subjects, across 13 studies, with symptoms of dry eye. For each subject the corneal zone of worst staining was recorded to analyse which region of the cornea most frequently exhibited the most severe staining. In Chapter 5, 20 subjects were exposed to a lens/solution combination, known to induce SICS, in both eyes for a two hour period. In phase one, one lens was rinsed thoroughly before being worn; in phase two, the eye itself was rinsed thoroughly post lens wear; in phase three, confocal microscopy was conducted on both eyes to look for hyper-reflective epithelial cells. In phases one and two, the epithelium was assessed for staining pre and post lens wear with and without fluorescein. Chapter 6 evaluated aspects of the staining data collected in several SICS-inducing studies. The frequency of the reported ‘donut’ pattern of staining was calculated, relative to a diffuse, pan-corneal staining pattern. Seven subjects were identified that had participated in three or more trials using the same SICS-inducing methodology. The data from these individuals were assessed to determine the repeatability of the level of induced corneal staining in these trials.

RESULTS The CORE corneal staining scale agreement experiment, in Chapter 3, supported the benefit of training because the concordance of the naïve observer was markedly worse than the observers who had received prior training. The inter- and intra-observer agreement analyses provided valuable data which can be applied to the development a pictorial reference guide and better instructions. The Chapter 4 meta-analysis of the geographic distribution of corneal staining among subjects with symptomatic dry eye demonstrated that the greatest degree of staining was most frequently in the inferior zone. (52.5%) The zone affected least was determined to be the central zone (12.8%). In the SICS experiment of Chapter 5, rinsing the lens prior to wear and rinsing the eye post lens wear did not result in different staining to the non-rinsed condition. All eyes, irrespective of any rinsing treatment, presented with punctate staining over >84% corneal area. The SICS staining was visible before fluorescein was instilled as ‘white light staining’. Confocal images were obtained from 34 of the 40 eyes, and hyper-reflective cells were visible in 33 of those 34 eyes. The meta-analysis in Chapter 6 concluded that the ‘donut-ring’ staining pattern, which is often described as typical of SICS, was actually far less common than a diffuse pan-corneal staining presentation. When SICS responding eyes were defined as exhibiting staining of 10% extent in at least four of the five corneal zones, 89% were identified as presenting with the pan-corneal pattern i.e. all five zones met the 10% extent criteria. When the SICS definition was tightened to include only those with 50% extent in at least four zones, 76% of subjects still identified as the pan-corneal staining pattern. There was minimal evidence of SICS presenting with a repeatable degree of staining in the same individual across different clinical trials.

CONCLUSIONS This thesis investigated several aspects of corneal epithelial fluorescein staining and the chapters have furthered understanding in this field in several ways. The CORE corneal staining scale provides valuable data regarding the percentage of the corneal affected by staining. The results of the Chapter 3 agreement experiment provide useful information for the next steps in the development of this scale which will create a valuable corneal staining assessment tool. The evidence that the most severe corneal staining in patients with symptoms of dry eye most often presents in the inferior zone is invaluable to the design of future clinical trials of dry eye treatments. It highlights the importance of specifically assessing this region and the value in targeting fluorescein staining improvements in this zone as a key outcome measure. SICS has been suggested to be due to adhesion between PHMB (or other care system components) and the epithelial cells. The experiment in Chapter 5 confirmed that rinsing the lens does not remove enough PHMB from the lens to prevent SICS, and rinsing the eye afterwards is not effective at removing the bound molecules from the epithelial cells because SICS is still evident post rinsing. The presence of ‘white light staining’ and hyper-reflective cells on in-vivo confocal microscopy indicate that there are changes to the epithelial cells even before the fluorescein is instilled into the eye. More investigation of changes at the cellular level are required to understand what is happening. The meta-analysis of SICS data was able to provide evidence that SICS most commonly presents as a diffuse punctate staining that affects the entire cornea presenting in a pan-corneal pattern, rather than presenting in the commonly described pattern of a donut-ring, which implies central zone sparing. The examination of SICS in seven subjects across several studies questions the repeatability of the SICS phenomenon. A targeted repeatability trial is required to conclusively answer this question.