From a functional perspective, the ideal ablation profile is not difficult to imagine: It would correct refractive and corneal higher-order aberrations with minimal biomechanical weakening of the cornea; avoid unwanted side effects, such as glare or halos; anticipate the cornea's healing response; resist epithelial remodeling, regression, and haze formation; and provide a refractive correction that covers the entrance pupil so that defocused light rays from the periphery of the cornea do not optically degrade the image. To accomplish this, the ideal ablation profile would adjust for the starting curvature and asphericity of each eye to deliver the optimal result.
In practice, ablation profiles are a compromise of optics versus physiology. Driving the desire for better ablation profiles is the goal of improving optical performance after surgery; delaying progress are the physiologic limitations of the cornea, variability among eyes, and technical limitations of lasers to deliver the desired treatment. Cost is also a factor because manufacturers weigh the costs of improving technologies that are already good enough against the uncertain potential for increasing profits.
With these points in mind, let us consider the features of the ideal ablation profile (Table 1).
OPTICAL FUNDAMENTALS
The cornea accounts for approximately 75% of the eye's refractive power. Therefore, the slightest change in corneal shape can greatly affect vision. The refractive elements of the cornea start with the tear film and include the anterior and posterior corneal surfaces. Images then pass through the aqueous and are affected by the lens before they travel through the vitreous to reach the retina. Every time incoming light passes from one refractive index to another—at each interface between media along this path—the image is affected.
Excimer-based refractive surgery procedures modify vision by changing the corneal shape. This is not the ideal approach.
Ideally, the cornea and crystalline lens would work together to create an undistorted image free of aberrations and focused precisely at the retina. To accomplish this, the cornea and lens would focus the image at exactly the axial length of the eye. In nonemmetropic eyes, this might require interventions at the cornea, the lens (which is currently unavailable to intervention), and even at the axial length of the eye (which has been tried, but with severe complications).1,2
The optical quality of cornea-based refractive procedures is limited by the distance of the cornea from the nodal point of the eye. Phakic IOLs are closer to the nodal point and can often provide better visual quality than cornea-based procedures; however, they introduce risks associated with intraocular procedures that are avoided by corneal procedures.
Guell, Pallikaris, Mrochen, and others have presented information at recent meetings suggesting that microscopic irregularities of the postablation stromal surface and the LASIK flap interface may lead to light scattering and diminished optical performance after LASIK. The ideal ablation profile has a perfectly smooth surface to minimize these effects.
CORRECTION OF LOWER- AND SELECTED HIGHER-ORDER ABERRATIONS
Despite our desire to deliver an aberration-free image to the retina, the optimal cornea-based procedure would not attempt to correct complex lens-based aberrations.
It has long been recognized that the crystalline lens can contribute to refractive errors,3 particularly astigmatism. As the nucleus hardens and ages, it develops an increasing amount of higher-order aberrations, particularly spherical aberration and coma.4,5 Refractive surgeons historically have incorporated lenticular astigmatism into cornea-based refractive corrections—despite the possibility of necessitating a refractive treatment when the natural lens is removed later in life. However, a recent US Food and Drug Administration (FDA) study showed that aberration-guided treatments (ie, wavefront-guided) on the cornea for asymmetric aberrations in the lens, such as coma, may increase the overall total-eye aberrations after surgery.6 The postoperative cornea bends the incoming light differently than the preoperative cornea did; therefore the light rays do not strike the lens at the same location as when the treatment was designed. The result can be increased, rather than decreased, aberrations. This does not generally occur with spherocylindrical corrections because the optics of spherocylindrical errors are mostly symmetric.
Although most surgeons are comfortable treating lens-based astigmatism with corneal corrections—even though those corrections may need to be reversed after subsequent cataract surgery—the treatment of lens-based higher-order aberrations on the cornea is not recommended. Such treatments may worsen vision and induce irregular corneal patterns that may be irreversible. Conversely, correcting cornea-based higher-order aberrations is often appropriate, provided that underlying disease processes, such as keratoconus or pellucid marginal degeneration, are not the cause of the aberrations. In this context, spherical aberration should be considered separately from coma or other higher-order aberrations.
What about spherical aberration? Corneas are naturally aspheric—the central curvature is different from the curvature in the mid-periphery—resulting in some degree of corneal spherical aberration that, in most eyes, is ideally counterbalanced by a reciprocal spherical aberration in the lens. If the corneal and lenticular spherical aberration do not reciprocally balance, then spherical aberration of the total ocular wavefront results. Usually, this occurs with early lens changes, and it should not be corrected because the changes are unstable. The issue is further complicated by radial asymmetry of corneal asphericity, which can lead to complex optics after treatment.
Cornea-based treatments of spherical aberration are as likely to increase spherical aberration as they are to decrease it. Previous studies7,8 suggest that the natural corneal asphericity of most eyes is optimal and should be left unchanged. Therefore, the ideal ablation profile corrects the refractive error but leaves corneal asphericity unchanged.
In simple terms, corneal asphericity is considered the ratio of corneal curvatures at two points. For refractive surgeons, curvatures outside the ablation zone are uncontrollable; the current approach is to use ablations that extend the refractive correction as far to the edge of the optical zone as possible. For example, an eye with a central curvature of 44.00 D, curvature at the optical zone's edge of 42.00 D, and a correction of 4.00 D would have corresponding approximate curvatures of 40.00 D and 38.00 D after surgery. This approach, leaving the ratio of central and mid-peripheral curvatures unchanged, is the underlying principle behind wavefront-optimized ablations.
Asymmetric corneal shapes can lead to coma and other aberrations. These can and should be treated because (1) there is no risk for an unpredictable summation of corneal and lenticular aberrations and (2) treatment will not complicate future cataract surgery. In today's setting, wavefront-guided treatments may be used for asymmetric corneal aberrations, particularly if the comparison of wavefront and topographic maps suggests that the lens is relatively symmetric; however, if there is evidence of significant lens aberrations, the best treatment might be a simple spherocylinder treatment, leaving the corneal aberrations untreated. In the future, topography-based treatments will permit direct treatment of cornea-based aberrations without introducing lens-based aberrations onto the cornea.
PUPIL SIZE CONSIDERATIONS
The entrance pupil accepts light from a large corneal area. Any defocused light rays that reach the retina result in a deterioration of visual quality. Because pupil size increases in mesopic light conditions, ablations with optical zones smaller than the entrance pupil risk causing glare and decreased mesopic visual function.
Additionally, patients with larger pupils are more symptomatic with small residual refractive errors.
Corneal thickness places limitations on ablation zone diameters. When the ablation size is approximately 6.5 mm or larger, the Stiles-Crawford effect and higher-level neural adaptation may make good vision in mesopic conditions possible for patients with pupil sizes larger than the optical zone.9 This does not mean that larger pupils do not result in decreased image quality—they do. Instead, it means that patients can adapt to the image and mitigate the effects of defocused, peripheral light rays.
As the ablation diameter increases, so does the required ablation depth for the refractive correction. The ideal corneal ablation would maximize the ablation area and avoid ectasia due to removal of excessive amounts of tissue. Pupil size changes with ambient light conditions. Recently, some laser companies have suggested the use of the 95% pupil,10 which is the maximal pupil diameter encountered in 95% of the patient's activities. Interestingly, because pupils may not be radially symmetric, some have suggested designing the shape of the ablation to match the physiologic pupil outline.10 Whether this concept proves workable or useful awaits clinical validation.
PHYSIOLOGIC LIMITATIONS
Corneal ectasia after excimer laser ablation concerns patients and surgeons alike. Convincing evidence11,12 suggests that most ectatic eyes were predisposed to keratoconus that may or may not have been identifiable prior to surgery. Nevertheless, as much stromal tissue as possible should be left behind, but still achieves the desired optical results. To that end, the ideal corneal ablation will extend as far to the periphery as possible and still respect the physiologic limits on ablation depth.
The recent surge in popularity of PRK is, in part, possible because of the smooth surfaces that result from ablations made with late-model excimer lasers and the low energy levels that flying spot lasers use to create the ablations.13,14 Nevertheless, excimer laser ablations that remove Bowman's layer face slower healing, greater postoperative refractive variability, refractive regression, and higher risk of haze formation that may occur even years after surgery.15 The use of mitomycin C has greatly decreased the incidence of haze; however, its long-term effects are unknown.
To avoid the complications of surface treatments, the ideal corneal ablation would be placed just beneath Bowman's layer, an approach called sub-Bowman's keratomileusis (SBK).16 The use of SBK is predicated on the ability to create a thin flap safely, and determining the size and optimal depth of those flaps is not trivial.17
CONCLUSION
Contrary to recent trends, the ideal ablation profile corrects manifest refractive errors and cornea-based aberrations but does not correct transient, unstable higher-order aberrations that originate from the crystalline lens. Instead, the ideal ablation profile leaves the cornea with undistorted surfaces so that subsequent lens surgery can be performed with a cornea that is free of irregular optics.
Because most eyes do not have symptomatic higher-order aberrations, the goal of corneal-based refractive surgery should be to correct refractive errors without inducing new aberrations. For most eyes, this is accomplished using wavefront-optimized ablations that preserve the cornea's preoperative asphericity. These ablations extend the effective optical zone further to the periphery so that the Stiles-Crawford effect minimizes mesopic visual symptoms in patients with wide pupils.
Current wavefront-optimized ablations are not perfect. They may still induce spherical aberration, particularly in eyes with extreme preoperative keratometry readings and unusual corneal asphericity. Corneal thickness limits the refractive range in which refractive errors can be treated without inducing spherical aberration to approximately 6.00 D using a 6.5-mm optical zone.18 Future technologies will allow surgeons to deliver wavefront-optimized optics to more eyes than current technologies allow.
Finally, the ideal ablation profile eliminates lower-order refractive errors and preserves biomechanical corneal integrity. Although the cornea is not the ideal place to correct refractive errors and cannot deliver optics that rival corrections performed at the nodal point of the eye, cornea-based refractive treatments can provide excellent vision for most patients. It is incumbent on the ophthalmologist and the ophthalmic industry to continue improving the ablation profiles that are available in our lasers and to strive toward these ideals in our quest to eliminate ametropia and to improve the visual function of our patients.
Guy M. Kezirian, MD, FACS, is President of SurgiVision Consultants, Inc., Scottsdale, Arizona. Dr. Kezirian states that he is or has been a consultant to Advanced Medical Optics, Inc., Alcon Laboratories, Inc., Bausch & Lomb, IntraLase, Inc., Visx, Inc., and WaveLight AG. He may be reached at e-mail: Guy1000@SurgiVision.net.