Research Review

Lissamine Green Versus Rose Bengal

Research Review

Lissamine Green Versus Rose Bengal


Ocular surface staining is one of the most widely used techniques in the clinical evaluation of dry eye patients. Historically, sodium fluorescein and rose bengal have been the two agents most commonly applied, but recent years have seen something of a stampede away from rose bengal and toward lissamine green. Without doubt, the view that this organic dye is easier on the eyes during instillation compared to rose bengal has been a significant factor in precipitating this migration. The evidence for such a comfort advantage does indeed exist (Maning et al, 1995; Amaki et al, 1999) and, of course, it can be easily verified by a little self-administration.

Over and above this, the question of whether these two stains can be regarded as interchangeable is an important one. Numerous studies have described the appearance of lissamine green staining in dry eye, and grading scales have been developed to assist clinicians in quantifying the severity of their observations; but do we really understand what these phenomena mean at the cellular level, and is this the same as would be the case had rose bengal been used instead?

Historical Findings

Although lissamine green had previously been used for staining the endothelia of eye bank eyes (Jans and Hassard, 1967; Kirk and Hassard, 1969), it was the great Danish ophthalmologist, Mogens Norn, who, some 40 years ago, was the first to advocate its use on the ocular surface (Norn, 1973). Based on a series of observations made across a range of pathologies, he concluded that lissamine green dye produced staining of degenerate and dead cells as well as staining of mucus.

Dr. Norn further noted that “its vital staining properties are almost identical with rose bengal.” This seems clear enough, but a re-reading of Dr. Norn’s original paper suggests that it might be worth taking a closer look at what he did.

Because Dr. Norn’s interest was in comparing staining effects on the same pathology and in the same eyes, he always used a mixture of dyes. Most often, this was lissamine green with rose bengal, but sometimes fluorescein was added as well, or the dyes were instilled sequentially, in various orders.

Dr. Norn cleverly employed a range of colored filters to distinguish the behavior of the dyes. But, with the benefit of hindsight, his interpretation of the observed staining may have been clouded by interactions between the various elements when simultaneously present in the eye. To give one example, both rose bengal and lissamine green are now known to have intrinsic toxicity toward epithelial cells (Chodosh et al, 1994), which means that what is revealed by one may be dictated or modified by the activity of the other. We will return to this in a moment.

Another feature of Dr. Norn’s study was that he extracted “mucus threads” from the subjects’ eyes and recorded that these were stained by the dye combinations with different colors—either red, green, or blue—occurring along the length of the thread. Presumably, this was the basis for his conclusion that lissamine green stains mucus.

Intriguingly, the staining of these strands was not always homogenous, with all three colors sometimes appearing at different points along their lengths. This behavior is strikingly complex, and it isn’t easy to work out how it can be explained in terms of simple staining of the mucus by components of the dye mixture.

Current Research Results

More information is needed to clarify this picture, and work conducted 20 years later, using cultured rabbit corneal epithelium cells, goes much of the way to doing that (Chodosh et al, 1994). Cell culture has big advantages in this area as it provides an opportunity to manipulate the various stains and observe effects on cellular metabolism that would be impossible in a clinical setting.

What this study showed was that lissamine green only stains epithelial cells if the cell membrane is damaged. It also showed that when this occurs, the dye concentrates mainly in the cell nucleus. This finding that lissamine green does not stain healthy cells has since been confirmed by other workers using human corneal tissue (Kim and Foulks, 1999) and establishes its status as a vital dye (i.e., it distinguishes between normal and damaged cells).

Note that the same cannot be said of rose bengal, because that substance actually produces staining irrespective of the state of cell health. Worse still, as mentioned earlier, it is actually cytotoxic (Feenstra and Tseng, 1992).

Both lissamine green and rose bengal are capable of killing cells on the ocular surface, but rose bengal is substantially more potent and faster acting. In fact, lissamine green is relatively benign unless exposure is prolonged beyond about 10 minutes (Chodosh et al, 1994; Kim and Foulks, 1999). Consequently, it can generally be regarded as safe, because a typical examination will be completed well within this time limit.

There were two other important outcomes from these cell culture experiments. The first is that unlike rose bengal, the presence of mucin on the surface did not prevent lissamine green from staining the cell, nor was the mucin itself stained in the process.

Although this second point is clearly contrary to Dr. Norn’s view, it does appear to be consistent with common clinical experience (i.e., it isn’t usual to see widespread lissamine green staining in normal eyes, despite the fact that mucin is found in many locations on the ocular surface).

The second useful feature is that lissamine green has an affinity for collagen. This, of course, means that we can expect it to highlight corneal stroma in the event that it becomes denuded of epithelium.

Conjunctival Considerations

It will not have escaped notice that the above discussion has been “cornea-centric.”And, while this is obviously an important site, the assessment of dry-eyed patients typically involves taking a broader perspective. The conjunctiva is often the focus of attention during clinical evaluation. So, it is surprising that this tissue hasn’t received much attention in the laboratory so far.

The single study that does exist (Doughty and Hagan, 2013) is an important one, however, because it nicely emphasizes the diverse behavior of our two chromatic assistants. Thus, while rose bengal vividly stains conjunctival epithelial cells, lissamine green does not. It is probably a reasonable presumption that staining will accompany damaged conjunctiva in the same way that it does corneal epithelium. But, this has yet to be confirmed experimentally.


Returning to the questions posed at the beginning of this article, we can sum up the state of our current knowledge by saying that lissamine green is a vital dye that stains membrane-damaged epithelial cells of the cornea (and probably conjunctiva) as well as exposed stromal tissue. It does not stain mucin, is relatively non-toxic, and causes little discomfort on instillation.

Rose bengal, on the other hand, is not a vital dye. However, in the absence of a protective mucin coating, it will stain both corneal and conjunctival epithelial cells, whatever their condition. It has intrinsic cytotoxicity and causes discomfort on instillation.

Clearly these two agents differ substantially in what they can tell us about the ocular surface. While we can agree that both have their place as clinical tools, and their use may even be complementary, they are certainly not interchangeable. CLS

For references, please visit and click on document #224.

Professor Papas is executive director of Research & Development, Brien Holden Vision Institute and Vision Cooperative Research Centre, and professorial visiting fellow, School of Optometry & Vision Science, University of New South Wales, Sydney, Australia. The Brien Holden Vision Institute and Vision Cooperative Research Centre have received research funds from B+L, AMO, and Allergan, and have proprietary interest in products from Alcon, CooperVision, and Carl Zeiss. You can reach him at