Corneal topography eases detection of early-stage keratoconus

Issue: August 1999

photograph ---Alcon’s EyeMap compares the level of asymmetry between the central cornea and the superior and inferior halves of the cornea, as well as the deviation of asymmetry from the pattern produced by a normal cornea.

While corneal topographers are helpful in assessing the shape of the cornea in postrefractive surgery and contact lens patients, practitioners say this technology is most useful for patients with unusual corneal conditions, such as keratoconus or corneal warpage.

In addition to evaluating the cornea to find the best lenses for keratoconus patients, certain topographers carry specific software for the purpose of detecting keratoconus in a patient. If this particular software is not available on a topographer, however, different aspects of the corneal map often will indicate whether or not the eye has become keratoconic.

Ken Daniels, OD, an assistant clinical professor and director of the Contact Lens Research Program at the Pennsylvania College of Optometry, said that he will perform topography on patients who exhibit a high level of astigmatism and are not correctable to 20/20. “My job is to ask, ‘Is this an astigmatic or meridional amblyopia, or is it an early form of a corneal problem?’” he said. “It helps you differentiate corneal vs. lenticular astigmatisms and assists in ruling out ocular media problems, such as form fruste keratoconus. These types of concerns aren’t as easily visualized when they are at subclinical phases. This is where corneal topography is of great clinical assistance.”

Dr. Daniels said that while it would be difficult to perform a topography exam on every patient due to the cost, the idea is a valid one. “It wouldn’t be a bad idea to do a baseline corneal topography on every patient, just like you do a baseline visual field,” he said. “You assume patients have glaucoma until proven otherwise; here you might want to assume they have keratoconus, or other corneal anomalies, until proven otherwise.”

Software detection systems

photograph ---Diagnosing keratoconus: Robert B. Mandell, OD, PhD, says a tangential corneal map, such as this produced by Dicon's CT 200, is better suited than an axial map to distinguish the difference in the maximum curvature of each of the patient's eyes.

Dean Burns, senior product manager for refractive and vitreoretinal products at Alcon, said that the EyeMap uses “an algorithm that takes into account what a normal cornea would look like in terms of the curvature,” he said. “Anything that deviates from what is designated as average, normal corneas, falls into an area of what we call suspect corneas. Then, those corneas get broken down into what we feel are true signs of keratoconus, early signs of the disease or something other than keratoconus. The software allows you to go from an objective state to something more defined. It’s either keratoconus or early signs of it, or a normal cornea with some signs of contact lens-induced warpage.”

The EyeMap uses the KI-SA percentage index, Mr. Burns explained. The calculation compares the level of asymmetry between the central cornea and the superior and inferior halves of the cornea, as well as the deviation of asymmetry from the pattern produced by a normal cornea. The percentage factor ranges from 0 to 60 for a normal cornea with a green color coding, 60 to 99 for a suspect cornea with a yellow coding and 100% and up in the red area for clinical keratoconus.

The Euclid system uses Fourier Profilometry, a software system that became available in August 1998, instead of placido rings. The system projects a series of parallel lines onto the eye to produce a map showing various heights of the corneal surface as opposed to curvatures. “In addition to several other maps, we take those depths and then fit an ellipsoid to the actual corneal surface, showing the differences in various colors,” said Randy DeWoolfson, vice president of systems at Euclid Systems. “In that way, any little bump, a cone in particular, pops off as a big red mark. You can measure its height at the apex and the diameter in any direction once you know where the boundaries are.”

Keratoconus-specific software

photograph ---Tomey’s TMS-3 includes software for keratoconus screening and contact lens fitting. The printouts for Tomey's systems will indicate "keratoconus detected."

Dr. Daniels said that he most often uses Tomey’s TMS-1 and TMS-2 topographers, which carry the Klyce/Maeda detection software and the Rabinowitz software, to test patients for the condition. “The Rabinowitz software does an inferior/superior comparison of symmetry to detect a dramatic difference that may indicate keratoconus,” he said. “Klyce/Maeda uses a whole series of qualitative descriptors separating the corneal map into a pie chart and then doing sector analysis to determine if there is the potential for or an indication of keratoconus.”

Dr. Daniels said that the system is not difficult to learn to use, and either the optometrist or a technician can perform the testing. He noted that while special software is not necessary in determining if a patient has keratoconus, the feature makes the condition easier to detect. “In the forme fruste keratoconus, which would be early formation without specific anatomy,” he said, “it definitely picks up on keratoconus you wouldn’t see otherwise. I’ve had at least 10 or 15 cases where the software picked it up that way, preventing me from proceeding with refractive surgery and assisting in guiding me to a more specific keratoconic contact lens fit.”

He said that the printout of the corneal map, a copy of which is given to the patient to take home, is also useful for insurance purposes. “With the Tomey system’s software, the printout says ‘keratoconus detected,’” he said. “We can print that and send it to the insurance company so there is irrefutable evidence that the patient has keratoconus. That helps tremendously in terms of practice management and helping the patient. The topographic printouts assist in educating to the patient and serve as a document for self-monitoring of the disease progression.”

The TMS-3, launched in November 1998, features a classification neural network trained to distinguish among different types of patterns of maps including clinical keratoconus, keratoconus suspect patterns, “normals,” astigmats, pellucid marginal degeneration, contact lens warpage, postradial keratotomy, postrefractive keratectomy and postpenetrating keratoplasty. When the Keratoconus Severity Index (KSI) reaches 0.15, the condition is considered suspect; for 0.30 and above, clinical keratoconus is indicated.

The automated software for keratoconus detection for Dicon’s corneal topographer, called the Bull’s Eye program, was developed in part by Robert B. Mandell, OD, PhD, a professor at the University of California School of Optometry in Berkeley. He said that by identifying a cone in any corneal position, as opposed to relying on a cone area that is decentered in an inferior direction as with some other systems, the Dicon system allows for earlier detection and a more detailed description of the cone shape for following the disease’s progression.

The system sends up a red flag when it encounters steepening and uses a bulls-eye targeting method that identifies the cone position and size on the corneal map in as little as 6 seconds, said Kathy Hand, marketing director at Dicon. “The system has an index that pops up on the screen to show the location of the elevation, the size and whether there’s a low, moderate, high or very high risk of keratoconus,” she said. “It indicates to the doctor that it is a suspect area. We also state that you should confirm the possibility of keratoconus through other methods as well, but at least this alerts you to the fact that there is an elevated, or peak, area on the cornea.”

Because superior flattening and inferior steepening of a cornea can suggest corneal warpage as well, software that analyzes the geometry of the cornea serves an important role in diagnosing keratoconus, said Kenneth Lebow, OD, who designed the Humphrey 991 system 2 years ago. The Pathfinder software helps differentiate contact lens-induced warpage from classic as well as subclinical keratoconus, he said, using overall corneal geometry including shape factor, regularity of the corneal surface and apical corneal toricity.

“The ‘shape factor’ is a relationship to corneal eccentricity,” said Dr. Lebow, in solo private practice in Virginia Beach, Va. “It deals with a rate of flattening of the shape of the eye. Typically, keratoconic eyes have larger, steeper shape factors, so they’re going to have relatively high eccentricities or shape factors. When you find the same topographic appearance with contact lens-induced warpage, you generally find a low to negative shape factor.”

Detect advanced cases without software

photograph ---Euclid’s ET-800 system uses Fourier Profilometry, a software system that involves a series of parallel lines shot onto the eye to produce a map showing various heights of the corneal surface as opposed to curvatures.

It is not necessary, however, for a corneal topographer to have special detection software to be an essential part of the keratoconus detection process, say practitioners. Dr. Mandell said that keratoconus, especially in the more advanced form, is easily detected with a topographer without such software.

“In more advanced keratoconus — depending on the position of the cone — just the general appearance of the corneal map is enough to tell you that a person has keratoconus,” he said. “The corneal map is particularly important as a detection method when the cone is in the central region of the cornea. The reason is that normal methods of detection, such as using a keratometer, often miss central keratoconus until it reaches fairly advanced stages.”

The base method to check for keratoconus, said Dr. Daniels, is simply to evaluate the map in terms of its symmetry. “The lesser the symmetry, the more indication there is of anomaly,” he said. “Then, the anomaly that you’re looking for in keratoconus is more of the centered, highlighted red area of the inferior portion of the map. The top part of the superior section looks flattened out. In the later stages, you don’t need software to detect that.”

Detection challenge: early keratoconus

photograph ---Pathfinder software on Humphrey's Atlas 991 helps practitioners determine whether cornea distortion is a result of keratoconus or contact lens-induced warpage by evaluating the overall geometry of the cornea.

When no special software is available on the corneal topography system, pinpointing keratoconus in its early stages using the map often presents a challenge. One of the identifying factors of the condition in its early stages is the area of maximum curvature, which falls beyond the normal range or below the center of the cornea, Dr. Mandell said. Because corneas differ greatly from person to person, it is difficult to differentiate between a healthy cornea and an abnormal one, he said; however, comparing each of the patient’s eyes to one another may help.

“There may be a significant difference between the maximum curvature of the two eyes, and that has to be determined in a certain way,” he said. “In many cases, it has been determined by using the axial, or basic, type of corneal map. This type of map has been grossly misused for this particular application because it has a significant error in determining results.”

Instead, he said, a tangential cornea map is better suited to distinguish the difference in the maximum value of the two eyes. “In keratoconus, one eye generally leads the other by a significant amount, and if there is a difference of at least 2 D, that’s an indication that keratoconus may be present,” he said.

Dr. Mandell recommended determining “the rate of change of the corneal power measured on the corneal topography map and going from the highest value at the apex to values that are adjacent to the apex, particularly the point that is 1.5 mm away from the apex,” he noted. “The reasoning here is that when you have keratoconus, the cone region of the cornea is bulging out, which results in a change in the slope of the adjacent area around the apex of the cone. That changes more quickly when you have a cone.”

Keratoconus vs. pellucid degeneration

Determining the difference between keratoconus and pellucid degeneration, a corneal thinning disorder that affects the interior cornea just above the limbus, is a good reason to use a topographer, said Arthur B. Epstein, OD, director of the contact lens service of North Shore University Hospital, New York University School of Medicine.

“This condition presents with a different pattern, typically with more astigmatism and less irregularity than keratoconus does,” he said, noting that pellucid degeneration patients often do well with soft lenses. “It’s important to make that differential diagnosis, because fitting a pellucid patient with a rigid lens, thinking it’s keratoconus without the benefit of topography, can be disastrous. We’ve seen several cases of well-intentioned colleagues who thought they were dealing with keratoconus fitting patients with rigid lenses and then having patients come back with half their corneas eroded over the thin area. From that point of view, it’s an important issue.”

Topography not foolproof

While the corneal topographer is an integral part of the keratoconus detection process, it should not be relied upon for a diagnosis, urged Dr. Daniels. He said that the technology can often be “fooled” by rigid gas-permeable patients with superior aperture fits. “If a patient has a superior aperture fit, it pushes down into the inferior cornea, and it can mimic the keratoconic eye,” he warned. “So you not only have to use your impression of the map and what the application tells you, but you have to go back and look at the clinical side of it. When I give lectures, I always say that if you base your plan on one test, you’re just going to hurt the patient as well as yourself. A topographer is really a tool to help you make the assessment, but it should not be solely relied upon for the assessment.”

Dr. Epstein agreed that the human factor is crucial in diagnosing the disease. “The diagnosis is still subject to human interpretation, no matter how you look at it,” he noted. “The algorithms used with the software and the common sense thinking and experience of experts applies, but it’s not the end-all, be-all of the diagnosis. That really relies on the interpretation of the topography.”

Even then, Dr. Epstein maintained, while using a topographer is definitely a fundamental part of managing the disease, the path to pursue is not always clearly indicated. “With all of the great technology we have and improvements in contact lenses, it still doesn’t substantially change the prognosis because we don’t really have an effective treatment,” he said.

For Your Information:

Ken Daniels, OD, is an assistant clinical professor and director of the Contact Lens Research Program at the Pennsylvania College of Optometry. He may be reached at 1200 West Godfrey Ave., MOD 6, Philadelphia, PA 19141-3399; (215) 276-6104. Dr. Daniels has no direct financial interest in the products mentioned in this article, nor is he a paid consultant for any companies mentioned.

Dean Burns is the senior product manager for refractive and vitreoretinal products at Alcon Laboratories Inc. He may be reached at 6201 South Freeway, Fort Worth, TX 76134; (817) 551-4726; fax: (817) 551-4030.

Randy DeWoolfson is the vice president of systems at Euclid Systems. He may be reached at 2810 Towerview Rd., Herndon, VA 20171; (800) 477-9396; fax: (703) 471-7577.

Robert B. Mandell, OD, PhD, is a professor at the University of California Graduate School, School of Optometry, Berkeley, CA 94720-2020; (510) 642-9679; fax: (510) 643-5109; e-mail: mandell@spectacle.berkeley.edu. Dr. Mandell has no direct financial interest in the products mentioned in this article. He is a paid consultant for Dicon.

Kathy Hand is the marketing director for Dicon. She may be reached at 10373 Roselle St., San Diego, CA 92121; (800) 426-0493, ext. 174; fax: (619) 554-0332; e-mail: khand@dicon.com.

Kenneth Lebow, OD, may be reached at 345 Edwin Dr., Virginia Beach, VA 23462; (757) 497-5555; fax: (757) 499-2636. Dr. Lebow has a direct financial interest in the Pathfinder software. He is a paid consultant for Humphrey Instruments.

Arthur B. Epstein, OD, is director of the contact lens service of North Shore University Hospital, New York University School of Medicine and is an adjunct assistant professor at the North Shore University College of Optometry. He can be reached at North Shore Contact Lens and Vision Consultants, P.C., 1025 Northern Blvd., Ste. 94, Roslyn, NY 11576; (516) 627-4090; fax: (516) 627-4169; e-mail: artepstein@ibm.net. Dr. Epstein has no direct financial interest in the products mentioned in this article, nor is he a paid consultant for any companies mentioned.

The EyeMap is available from Alcon Laboratories, 6201 South Freeway, Fort Worth, TX 76134; (800) 862-5266; fax: (800) 241-0677; Web site: www.alconlabs.com.

The ET-800 is available from Euclid Systems, 2810 Towerview Rd., Herndon, VA 20171; (888) 338-2543; fax: (973) 773-6334; e-mail: infor@euclidsys.com.

The TMS-1, TMS-2 and TMS-3 are available from Tomey Corp. USA, 300 Second Ave., Waltham, MA 02154; (781) 890-1515; (800) 358-6639; fax: (781) 290-5885; Web site: www.tomey.com.

The CT 200 is available from Dicon, 10373 Roselle St., San Diego, CA 92121; (619) 554-1770; (800) 426-0493; fax: (619) 554-0332.

The Atlas 991 is available from Humphrey Systems, 5160 Hacienda Dr., Dublin, CA 94568; (800) 423-4393; fax: (925) 557-8025; e-mail: info@humphrey.com, Web site: www.humphrey.com.