Article Date: 6/1/2010

Exploring Beyond the Corneal Borders

Exploring Beyond the Corneal Borders

Not much is known about the shape of the corneal-scleral transition area and the anterior sclera.

By Eef van der Worp, BOptom, PhD, FIACLE, FAAO, FBCLA; Tina Graf, BSc; & Patrick J. Caroline, FAAO

Dr. van der Worp received his optometry degree from the Hogeschool van Utrecht in the Netherlands. He lectures extensively worldwide and he resides both in Amsterdam and in Washington, DC.

After finishing her optics degree in Germany, Tina Graf worked for two years as a contact lens specialist. For her optometry degree at Aalen University in Germany she completed a project at Pacific University on anterior ocular surface shape.

Patrick Caroline is an associate professor of optometry at Pacific University. He is also a consultant to Paragon Vision Sciences.

Large-diameter contact lenses that have their resting point at least partly beyond the corneal borders are believed to be among the best vision correction options for irregular corneas; they can postpone or even prevent surgical intervention as well as decrease the risk of corneal scarring. But not much is known about the shape of the corneal-scleral transition area and the anterior sclera. This article explores the shape of the anterior ocular surface beyond the corneal borders.

Indications and Options

Indications for scleral lens fitting have been evolving over the last few years, emerging from a lens for severely irregular corneas only to a much broader spectrum of indications. These include moderately irregular corneas, primary corneal ectasia (such as keratoconus, keratoglobus, and pellucid marginal degeneration), and secondary ectasia (post-refractive surgery), but also as a means of corneal protection such as in persistent epithelial corneal defects and conditions such as Stevens-Johnson Syndrome, graft versus host disease, ocular cicatricial pemphigoid, exposure keratitis/ocular surface disease, neurotrophic corneal disease, and atopic keratoconjunctivitis.

Corneal ectasia, including keratoconus, is still the main indication for fitting scleral contact lenses to restore vision. The Keratoconus Foundation estimates that about 15 percent to 20 percent of these patients will eventually undergo surgical treatment for the condition. The main form of surgical intervention in keratoconus is a keratoplasty. Keep in mind, though, that the survival rate of penetrating corneal grafts is 74 percent after five years, 64 percent after 10 years, 27 percent after 20 years and is very limited at 2 percent after 30 years (Borderie et al, 2009). Partial keratoplasties (lamellar keratoplasty) in which only the anterior portion of the cornea is removed can somewhat help overcome these problems, mostly because of the lower risk of graft rejection. They also may result in better visual outcome in the hands of a skilled surgeon.

But even when medically successful and without complications, many patients post-keratoplasty still need a contact lens, usually a GP lens, to restore vision because of irregularities and high corneal astigmatism. It is estimated that 90 percent of corneal ectasia patients will need GP lenses at some point to achieve acceptable vision. A study by Smiddy et al (1988) found that 69 percent of patients who were referred for a keratoplasty could be successfully fit with contact lenses without surgery. These statements seem to indicate a need for eyecare practitioners to evaluate all contact lens options first before referring a patient for surgery.

In the Collaborative Longitudinal Evaluation of Keratoconus (CLEK) study in the United States, 1,209 keratoconus patients were observed over a period of eight years at different sites. Results from the CLEK study show that scar formation in keratoconus may lead to a loss in contrast sensitivity, which may create a vision problem (Barr et al, 1999). This is especially a concern because keratoconus patients already have increased higher-order aberrations, primarily vertical coma, that may result in reduced contrast sensitivity. Baseline factors predictive of incident scarring included corneal curvature greater than 52.00D, contact lens wear, marked corneal staining, and a patient age of less than 20 years. Avoiding pressure on the apex of the cornea with contact lenses seems advised.

Different fitting approaches with GP lenses that aim for this fitting relationship are available. GP lens fitting has evolved and improved dramatically over the last 10 years with sophisticated lens designs based on corneal topography (such as highly aspheric and quadrant-specific lens designs). But despite this, reducing mechanical stress on the cornea is a challenge with every keratoconus lens. If a corneal lens does not work for these corneas, a scleral lens may be necessary to restore vision to help reduce the potential for increased scarring.

Corneal Clearance

For true clearance of the cornea, without any mechanical involvement, it seems advised to avoid any contact between the lens and the cornea by bridging over it. Large diameter contact lenses that have at least part of their resting point beyond the corneal borders are believed to be among the best vision correction options for irregular corneas. They can postpone or even prevent surgical intervention as well as decrease the risk of corneal scarring.

A standard definition for this category is not agreed upon, but generally corneo-scleral or semiscleral lenses partially rest on the cornea/limbus and partially on the anterior sclera. Typically they are in the range of 12.50mm to 15.0mm in diameter, and they have a limited tear reservoir (clearance) underneath of the lens. Scleral lenses are larger, ranging from 15.00mm to 24.00mm in diameter, and rest entirely on the sclera. Lenses that exclusively rest on the sclera can be further subdivided into large and small scleral lenses.

The biggest difference between the small- and the large-diameter scleral lenses is the amount of clearance present underneath of the central lens. In smalldiameter lenses the tear reservoir capacity is limited, while in the large-diameter scleral lenses the tear reservoir capacity is almost unlimited. The cut-off between these two is somewhat arbitrary, but most people would agree that a full-scleral, large-diameter lens would fall in the 18.00mm to 24.00mm range while the small scleral lenses, sometimes referred to as mini-scleral lenses, would be in the 15.00mm to 18.00mm range. Practitioners in the United States mostly use the corneo-scleral type lenses, while quite a few specialty practices in Europe almost exclusively use full-scleral contact lenses. All types of (semi-) scleral contact lens designs have the ability to promote good apical clearance as opposed to corneal contact lenses, which can reduce mechanical stress to the cornea and is the major advantage of this contact lens type.

But not much is known about the shape of the corneal-scleral transition area (referred to as limbal shape here) and the anterior sclera, in contrast to the vast amount of knowledge we have about corneal shape. What do we know about scleral anatomy?

Old School

First of all, textbook knowledge tells us that when looking at the insertion of the eye muscles on the eyeball, it appears that in the temporal, superior, and inferior direction there is roughly 7.00mm of space between the limbus and the insertion of the muscle itself (7.00mm, 7.50mm, and 6.50mm respectively). However, on the nasal side there is only a 5.00mm space available. With an average corneal diameter of 12.00mm, this means that horizontally 24.00mm is the maximum average physical diameter a contact lens can have before it may interfere with the eye muscle insertion.

Because of the anatomical location of the eyeball in the orbita, the temporal eye muscle wraps around the globe and stays in contact with it at all times, regardless of whatever eye movement occurs. The nasal eye muscle, on the other hand, comes loose of the globe with a medial eye movement despite its more anterior position of insertion on the eyeball. In a chapter of the book Contact Lenses, scleral lens expert Ken Pullum, FCOptom, DipCLP, describes in a section on scleral lens fitting that, "with large diameter lenses, this could theoretically mean that a lateral movement of the lens on the eye or a slight lift of the lens off the cornea can occur." Furthermore, he describes that it appears that the limbus on the temporal side of the cornea is less pronounced on average than it is on the nasal side because the center of curvature of the temporal sclera curve is contralaterally offset. Basically this means that the nasal scleral portion appears "flatter." In addition, the nasal curve of the sclera is often actually flatter, adding to the effect of a flatter nasal than temporal portion of the sclera (Pullum, 2005).

The shape of the limbus and the first part of the sclera beyond the limbus is subject to discussion. Previously it has always been assumed that the sclera has a curved shape, but this is not necessarily always the case. From the molds taken of the anterior segment of human eyes (in normal eyes and in keratoconus) it appears that at least in some cases, from the peripheral cornea onward the sclera often continues in a straight line (tangential). Also, when using contour maps from the experimental Maastricht Shape Topographer, one of the first topographers to image the limbus and part of the sclera up to 18.00mm in diameter of the anterior eye surface, it seems on a case-by-case analysis that the transition is often tangential rather than curved (Figure 1).

Figure 1. Anterior ocular surface contour map using the Maastricht Shape topographer (up to 18mm). Courtesy of John de Brabander

It is surprising how little we know about the limbal shape, which is a very important parameter when fitting contact lenses—soft lenses and of course scleral lenses in particular. One of the few publications on this topic can be found in the German contact lens literature. Daniel Meier, a Swiss eyecare practitioner, defines in die Kontaktlinse (October 1992) different transition profiles from cornea to sclera (Figure 2). He describes five different models: a gradual transition from cornea to sclera, where the scleral portion is either convex (profile 1) or tangential (profile 2), or a marked transition where again the scleral portion can be either convex (profile 3) or tangential (profile 4). As a fifth option, he describes a convex corneal shape with a concave scleral shape (profile 5). The profiles in the Meier scale are decreasing in sagittal depth, in which profile number 1 has the highest sagittal height and profile number 5 has the lowest—an important parameter for fitting scleral lenses.

Figure 2. Different transition profiles from cornea to sclera. Courtesy of Daniel Meier/die Kontaktlinse

Studies by Meier, and a study published in die Kontaktlinse (May 2001) by Rott-Muff et al, tried to identify how often the different profiles were seen in the general population. The study results were remarkably similar. Profile 2 (gradual-tangential) followed by profile 3 (marked-convex) were respectively the number one and two in presence, followed by profile 1 (gradual-convex). Profiles 4 and 5, marked-tangential and convex-concave, were seen minimally, with the latter one almost nonexisting.

The next question is: how accurately can these profiles be subjectively rated by practitioners? This also was addressed in an article in die Kontaktlinse a few years later (2007). Stefan Bandlitz and his students at the Cologne school of optometry investigated this, and found a repeatability of only 54 percent using 73 investigators. For some profiles the repeatability was much lower.

New Insights

In our studies, we wanted to repeat this experiment with the newest technology, but decided to simplify the method because of the large spread of outcomes. The criteria for describing the transition from cornea to sclera (limbal shape) was simply to mark it as either gradual or marked. We used optical coherence tomography (OCT) to take pictures of eyes in the large scleral lens practice of Greg Gemoules, OD, in Dallas, Texas. We used applied software to manually draw a forced circle through the periphery of the cornea, and a circle through the anterior sclera (Figure 3).

Figure 3. Examples of a gradual transition from cornea to sclera (left) and a marked transition from cornea to sclera (right).

The results of the 46 analyzed profiles were that the average peripheral corneal radius was 9.10mm (range 7.80mm to 10.80mm) and the average anterior scleral radius was 12.40mm (range 10.10mm to 16.60mm). Note that some peripheral corneal radii were flatter than were some anterior corneas. The median difference between the two was 3.40mm (range 1.50mm to 6.50mm) in radius, which we used as the critical cut-off point for gradual versus marked transition. Using this criteria, the distribution was 50-50 for abrupt versus marked. We also had, in a masked fashion, three different investigators observe and rate the 46 limbal profiles subjectively. In 75 percent of cases the subjective observation by the masked investigators correlated with the objective measurement by the computerized method. In 70 percent of the cases, the observers agreed with each other on the type of profile. The distribution for the subjective grading by the investigators also was roughly 50-50 for marked versus gradual.

We decided to take this one step further, again using anterior segment OCT. In our Pacific University College of Optometry studies, the corneal-scleral tangential angle between 10.00mm and 15.00mm (defined in our study as the limbal angle) was measured as well as the angle from 15.00mm to 20.00mm (the scleral angle) (Figure 4).

Figure 4. Limbal and scleral angle position illustration with the OCT in the nasal-inferior direction of an left eye in a study subject.

First of all we wanted to know: what is the shape of both the limbal area and the anterior scleral area—convex, concave, or straight? We took measurements of 96 eyes of 48 normal subjects in eight different directions: (nasal, nasal-inferior, inferior, inferiortemporal, temporal, temporal-superior, superior, and superior-nasal). Purely on theoretical considerations, we would expect the limbal area to be concave. But contrary to that general belief, the shape of the transition area between cornea and sclera was straight in most cases, with only one-quarter of cases exhibiting concave shapes as well as few convex cases. In addition, illustrating the individual character of the limbal shape, within one eye different profiles were measured in different meridians. So what about the anterior scleral shape (between 15.00mm and 20.00mm diameters)? In this case, again based on theoretical considerations, we would anticipate the anterior scleral shape to be convex: the eye is an eyeball in the end. But instead we found that in most cases the anterior scleral shape was tangential, with the expected convex shape a distant second in roughly less than one-third of the cases, and a minimal number of concave shapes.

To cut things short from a practical viewpoint: don't expect the limbal area and the anterior sclera to necessarily have the concave/convex shapes you would expect based on theoretical consideration when fitting/designing a scleral lens. Based on this, it is suggested that using tangent angles rather than using curves (or very flat curves) may be most appropriate in the majority of cases when fitting scleral lenses (van der Worp and Caroline, 2009) However, differences in individual limbal and anterior scleral shape exist, even within the same eye in different meridians.

Following this, we took all measurements together (1,289 angles could be measured in total) to form an anterior-shape-model-eye. The biggest surprise in this case was that the ocular surface beyond the cornea is not rotationally symmetric, illustrated in figures 5a and 5b. Typically, in the average eye the nasal portion is flatter compared to the rest, which is in line with corneal topography findings; the peripheral cornea is typically flatter nasally. Interestingly, the nasal scleral angles are typically flatter compared to the limbal scleral angles, which means that the sclera does not steepen as you would expect, but flattens out. In the other directions, the scleral angle was typically steeper compared to the limbal angle. The inferior segment of the eye typically is ‘on par’ both for the limbal and for the scleral angle, with almost no difference between the two. The temporal portion of the anterior ocular surface typically is steeper compared to the other areas—for example, the angles are higher in value. Superiorly is somewhat in between nasal and temporal, but with a substantial difference between the limbal and scleral angle. See Figure 6 for an example of a subject who has a steep overall appearance.

Figure 5a and 5b. Examples of the right eye of two normal subjects with their limbal and scleral angles in eight directions with a very flat appearance (PG) and a very steep appearance (TS), both with their corneal topography images superimposed.

Figure 6. Example of a tangential limbal and anterior scleral shape OCT image, with a 41 degree limbal and scleral angle (steep profile) in the 225 degree direction.

Both the limbus and the anterior sclera show the same pattern, but in the limbal area (10.00mm to 15.00mm) these differences are relatively small, although with 100 microns in magnitude these differences within one eye could still have clinical significance. For the scleral shape (15.00mm to 20.00mm) though, the differences are in the 400 micron range within one eye and surely are clinically relevant.

Summarizing this part of the study: the average eye is nonrotationally symmetric beyond the corneal borders. It appears that for an average eye, nonrotationally symmetric lenses such as toric and quadrant-specific lenses, both which are commercially available, could be the preferred option to optimally respect the shape of the eye. This is especially the case if the lens fit goes beyond 15.00mm, which is common in many parts of the world.

The nonspherical nature of the sclera has been described previously by Visser et al (2006) in a paper on toric scleral lens designs. In fact, more often than not, toric designs are used today when fitting scleral lenses in the Visser practices in the Netherlands. These nonrotationally symmetric lenses can lead to better ocular health because fewer areas of pressure are created, which can lead to a reduction in local conjunctival blanching—a term used to describe a decrease in local conjunctival blood supply.


More often than not, the shape of the limbus and anterior sclera appears to be tangential in shape, which has clinical consequences when fitting and designing scleral lenses. The fact that the sclera is nonrotationally symmetric in nature seems to suggest that toric and/or quadrant-specific scleral designs could lead to better fitting characteristics and possibly better lens comfort. But furthermore, because these lenses follow the shape of the anterior eye beyond the cornea more precisely, they are exceptionally stable on the eye, which opens up the possibility for additional optical corrections such as front-cylinders and higher-order corrected aberrations such as vertical coma, a very frequent finding in keratoconus. This can also help improve visual performance, which can further benefit patients who have keratoconus and other corneal irregularities. CLS

Special thanks to the Pacific University College of Optometry, especially Beth Kinoshita, OD, FAAO, and Matt Lampa, OD, FAAO, for their contributions to the OCT studies (Pacific University) and Greg Gemoules, OD, for his hospitality and help with the corneal-scleral transition study.

For references, please visit and click on document #175.

Contact Lens Spectrum, Issue: June 2010