A Standardized Classification of Corneal Topography after Laser Refractive Surgery

Peter S. Hersh, M.D.

From Journal of Refractive Surgery 1997;13:571-575

From the Department of Ophthalmology, UMDNJ-New Jersey Medical School, Newark, New Jersey; the Cornea and Laser Institute, Hackensack University Medical Center, Teaneck, New Jersey.

This study was supported in part by an unrestricted grant to the Department of Ophthalmology from Research to Prevent Blindness, Inc., New York.

With the expanding literature and discourse at professional meetings regarding the influence of corneal topography on outcomes following laser refractive surgery, the need for a universal, standardized classification system has become increasing apparent. We have previously reported a 7-category topography classification schema for eyes following photorefractive keratectomy (1,2). In this paper, we propose this original system with minor changes as a standard for use when evaluating corneal topography after laser refractive surgery.

The rigorous classification of topography patterns is important for a number of reasons: (1) To facilitate comparison of clinical studies and clarify professional communication. For instance, is a "central island" in Study A designated a "keyhole" in Study B? (2) To better understand the source of postoperative optical aberrations such as glare, halo, and monocular multiplopia. Do specific topography patterns cause specific optical side effects? (3) To facilitate studies focusing on the potentially unique etiologies of different patterns. Are hydration shifts, beam inhomogeneity, plume effects, wound healing, and other, as yet, undiscovered causes related to specific topography patterns? Are some patterns elevations whereas others are depressions? (4) To allow for natural history studies of topography patterns after laser procedures. Is the keyhole pattern a resolving central island, a forme fruste of the latter, or unrelated? Will visually significant patterns resolve on their own? (5) To suggest initial surgical strategies to avoid less desirable patterns. Should treatments be continuous or interrupted? Should the cornea be dried? Are pretreatments, multizone, multipass, or other techniques advisable? (6) To suggest postoperative medical and surgical treatment interventions. What patterns should be treated? How should they be treated? When should intervention be considered?

These and other such questions remain, as yet, unanswered and call for rigorous clinical investigation. An appropriate topography classification schema would thus comprise distinct and clinically useful patterns to maximize our ability to address these issues.

A number of investigators have already identified a variety of patterns following excimer laser photorefractive keratectomy. Lin and coworkers, using differential maps, originally classified topography into one of four patterns -- central uniform, keyhole, semicircular, and central bump (known in the literature afterwards as a central island) (3,4). In their system, the semicircular pattern comprised a non-uniform crescentic pattern with a power difference of 1.50 D or more across any meridian, extending for 90 degrees or more. The keyhole pattern was defined as a non-uniform pattern with a power difference of 1.50 D or more between equidistant points on any meridian relative to the center of the ablation zone and extending less than 90 degrees. They defined a central island as an increase in the central power of the ablation zone of 1.50 diopter or more and occupying 2.50 mm or more. Krueger and colleagues, using postoperative topography maps, defined the central island as topographic areas of steepening of at least 3 D and 1.5 mm in diameter (5). In contrast, Levin and coworkers, using 3 month postoperative topography maps, defined a central island as any part of the treatment zone surrounded by areas of lesser curvature on more than 50% of its boundary and differentiated between central islands of different heights and diameters (6). This definition actually includes the keyhole pattern as described by some other investigators.

Despite these excellent studies, the refractive surgery literature remains confusing and inconsistent in its topography taxonomy. This makes comparison of studies difficult and obfuscates communication amongst clinicians. In general clinical discourse today, for instance, the term "central island" is often used to describe a wide range of irregular topography patterns which potentially may have distinct etiologies or clinical sequelae. Communication in our field, therefore, would be facilitated by a universal classification system for topography patterns. Most work in laser refractive surgery, to date, has been done with spherical photorefractive keratectomy (PRK). However, a similar schema would be useful in future investigations of astigmatic PRK and the laser in situ keratomileusis (LASIK) procedure.

Given the requirements of clinical and research efforts, we propose an 8-category standardized classification system of corneal topography after laser refractive surgery, and encourage its use to facilitate research and clinical communication in the future.

Methodology

We base our topography pattern classification on the differential topography map, derived from subtraction of the preoperative from postoperative values at corresponding points on the actual power maps. The differential map, in theory, is preferred since it reflects the actual changes induced by the laser treatment and subsequent wound healing, accounting for pre-existing topography features unique to a specific patient. More than one differential map should be reviewed at each examination since these may differ between examinations even on the same day (7).

Color bins of 0.5 diopter are used for increased sensitivity in identifying possibly significant patterns (8). Furthermore, the actual videokeratograph should also be reviewed, since frank surface irregularities may be misrepresented by the topography system's reconstruction algorithm as a specific topography pattern rather than as an optically undefinable area (2). The measurement instrumentation and software version should always be noted since different systems and reconstruction algorithms may affect the postoperative topography pattern (9-13).

Standardized Classification System for Corneal Topography

Patterns after Laser Refractive Surgery

1. Homogeneous (Figure 1A) - shows a uniform and symmetric flattening, often with a smooth power change gradient with a progressively decreasing power change proceeding from the center to the periphery of the treatment zone.

2. Toric-with-axis (Figure 1B) - shows a meridionally symmetric treatment zone with a greater induced flattening (toric bowtie) in the steep preoperative axis. The differential between flattening and steepening in the two orthogonal meridians should be >0.5 diopters. This pattern would presumably decrease the patient's corneal toricity. Moreover, it is the expected pattern resulting from astigmatic PRK.

3. Toric-against-axis (Figure 1C) - shows a meridionally symmetric treatment zone with a greater induced flattening (toric bowtie) in the flat preoperative axis. The differential between flattening and steepening in the two orthogonal meridians should be >0.5 diopters. This pattern would presumably increase the patient's corneal toricity.

4. Semicircular (Figure 1D) - shows a general foreshortening of the treatment zone effect in one meridian, quantitatively measuring >1.0 mm in size and >1.0 diopter of relatively less flattening compared with the meridian 180 degrees away.

5. Keyhole (Figure 1E) - shows topographic regions, quantitatively measuring >1.0 mm in size and 1.0 diopter of relatively less flattening, extending in from the periphery of the ablation zone in one meridian. Unlike the semicircular pattern, the area of relatively less power decrease encompasses less than 180 degrees of the periphery and extends as an apparent involution of diminished power decrease into the treatment zone.

6. Central island (Figure 1F) - shows a central area of relatively less flattening measuring >1.0 mm in diameter and >1.0 diopter in power, and not extending to the periphery. The central island is surrounded around its entire periphery by areas of greater power diminution.

7. Irregularly irregular (Figure 1G) - shows generalized irregularities over the treatment zone, defined as more than one area measuring >0.5 mm and >0.5 D in power from other areas at the same radius from the treatment zone center, and not conforming to the specific criteria of the other patterns described.

8. Focal topography variant (Figure 1H) - shows a generally homogeneous pattern with a single irregularity measuring <1.0 mm in size or <1.0 diopter in power. Although this is only slightly different from the homogeneous pattern, it is intended to determine if minor topography variations, not meeting the criteria of the other patterns, affect clinical outcomes. This pattern has not been found to have clinical implications in past investigations (1,2,14,15).

In addition to the eight patterns described, the topography subgroups may be collapsed into 2 broader categories for greater power on statistical analysis in research studies -- for instance, the homogeneous, toric-with-axis, toric-against-axis, and focal topography variant combined into one broader regular group and the keyhole, semicircular, central island, and irregularly irregular combined into one broader irregular group. In previous work, we have validated the selection of these pooled patterns as "regular" or "irregular" using a quantitative descriptor of corneal topography (14).

Discussion

The ultimate goal of refractive surgery is the production of a corneal surface which optimizes both visual acuity and other objective measurements of visual function while minimizing aberrations which would degrade the retinal image. Indeed, optical side effects such as glare, halo, and multiplopia may prove to be amongst the most exasperating to both patient and surgeon. A methodical understanding of corneal topography following excimer laser procedures is thus essential to meet the dual goals of explaining today's clinical results and improving the overall visual outcome of the procedure in the future.

A standardized classification scheme for corneal topography after laser refractive surgery should afford clinicians a common language in communicating topography features of their postoperative patients, allow a greater understanding and analysis of topography in the future, and facilitate comparison of clinical studies. "Lumping" and "splitting" in a standardized classification system should be directed at defining potentially distinct patterns from the point of view of both etiology and clinical consequence. Thus, as we learn more, additional patterns may be defined or two or more patterns may be determined to be equivalent in their etiology and clinical effect.

In our past work (1,2) and in other studies (3), the differential topography map has been used. In theory, this map is preferred since it reflects the actual changes induced by the laser treatment and subsequent wound healing, accounting for pre-existing topography features (eg, astigmatism). However, the reproducibility of the differential map has been questioned; one group has found that the differential map may differ amongst between examinations even on the same day (7). As yet, there is no literature on the reproducibility of postoperative topography pattern categorization using postoperative maps. Thus given this information, it likely is advisable that more than one differential map be reviewed at each examination before ascribing a particular category to a patient. Moreover, the postoperative map should also be considered, especially when planning a specific treatment strategy based on topography findings.

Further clinical and basic research in the field, certainly, is required to define the optimum methodology of corneal topography analysis. Until that time, the schema described may be a useful guide for ophthalmologists and researchers alike.

References

1. Hersh PS, Schwartz-Goldstein B. Summit PRK Topography Study Group. Corneal topography of phase III excimer laser photorefractive keratectomy: Characterization and clinical effects. Ophthalmology 1995;102:963-978.

2. Hersh PS, Shah S. Corneal topography of excimer laser photorefractive keratectomy using a 6-mm beam diameter. Ophthalmology 1997 (in press).

3. Lin DTC, Sutton HF, Berman M: Corneal topography following excimer photorefractive keratectomy for myopia. J Cataract Refract Surg 1993;19:149-54.

4. Lin DTC. Corneal topographic analysis after excimer laser photorefractive keratectomy. Ophthalmology 1994;101:1432-1439.

5. Krueger RR, Saedy NF, McDonnell PJ. Clinical analysis of steep central island after excimer laser photorefractive keratectomy. Arch Ophthalmol. 1996;114:377-381.

6. Levin S, Carson CA, Garrett SK, Taylor HR. Prevalence of central islands after excimer laser refractive surgery. J Cataract Refract Surg. 1995;21:21-26.

7. Johnson DA, Haight DH, Kelly SE. Reproducibility of videokeratographic digital subtraction maps after excimer laser photorefractive keratectomy. Ophthalmology 1996;103:1392-1398.

8. Wilson SE, Klyce SD, Husseini ZM. Standardized color-coded maps for corneal topography. Ophthalmology 1993;100:1723-727.

9. Roberts C. Characterization of the inherent error in a spherically-biased corneal topography system in mapping a radially aspheric surface. J Refract Corneal Surg 1994;10:103-21.

10. Mandell RB. The enigma of the corneal contour. CLAO J 1992;18:267-273.

11. Maguire LJ, Wilson SE, Camp JJ, Verity S: Evaluating the Reproducibility of Topography Systems on Spherical Surfaces. Arch Ophthalmol. 1993;111:259-262.

12. Mandell RB. Corneal power correction factor for photorefractive keratectomy. J Refract Corneal Surg 1994;10:125-8.

13. Roberts C. The accuracy of 'power' maps to display curvature data in corneal topography systems. Invest Ophthalmol Vis Sci 1994;35:3524-32.

14. Hersh PS, Shah S, Geiger D, Holladay J. Corneal optical irregularity following excimer laser photorefractive keratectomy. J Cat Refract Surg. 1996;22:197-204.

15. Hersh PS, Shah S, Holladay JT. Corneal asphericity following excimer laser photorefractive keratectomy. Ophth Surg Lasers 1996;27S:421-428.

Figure Legends

Figure 1: Topography classification schema. Differential maps derived by subtracting preoperative from postoperative corneal powers are shown. (A) Homogeneous - shows a uniform and symmetric flattening often with a smooth power change gradient with a progressively decreasing power change proceeding from the center to the periphery of the treatment zone. (B) Toric-with-axis - shows a meridionally symmetric treatment zone with a greater induced flattening (toric bowtie) in the steep preoperative axis. The differential between flattening and steepening in the two orthogonal meridians should be >0.5 diopters. (C) Toric-against-axis - shows a meridionally symmetric treatment zone with a greater induced flattening (toric bowtie) in the flat preoperative axis. The differential between flattening and steepening in the two orthogonal meridians should be >0.5 diopters.

(D) Semicircular - shows a general foreshortening of the ablation zone effect in one meridian, quantitatively measuring >1.0 mm in size and 1.0 diopter of relatively less flattening compared with the meridian 180 degrees away. (E) Keyhole - shows topographic regions, quantitatively measuring >1.0 mm in size and 1.0 diopter of relatively less flattening, extending in from the periphery of the ablation zone in one meridian. The area of relatively less power decrease encompasses less than 180 degrees of the periphery and extends as an apparent involution of diminished power decrease into the treatment zone. (F) Central island - shows a central area of relatively less flattening measuring >1.0 mm in diameter and >1.0 diopter in power, and not extending to the periphery. The central island is surrounded around its entire periphery by areas of greater power diminution. (G) Irregularly irregular - shows generalized irregularities over the treatment zone, defined as more than one area measuring >0.5 mm and >0.5 D in power from other areas at the same radius from the treatment zone center, and not conforming to the specific criteria of the other patterns described. (H) Focal topography variant - shows a generally homogeneous pattern with an irregularity measuring <1.0 mm in size or <1.0 diopter in power.