Executive Summary
In October 1997 the American Institute of Biological Sciences convened the Peer Review Panel on Photorefractive Keratectomy (PRK) Research. The panel conducted an independent review of the research literature concerning the procedure known as photorefractive keratectomy. The panel reviewed the available literature and
- evaluated the quality and merit of the scientific and clinical data;
- assessed the immediate and long-term outcomes of PRK;
- related the information, as much as possible, to visual performance specific to military tasks;
- outlined what is known and is not known about the procedure;
- outlined knowledge gaps in the scientific and clinical literature; and
- made recommendations for additional studies, some of which can be conducted only in the military setting.
PRK is a procedure that is gaining acceptance in the civilian population. It offers the myopic individual the hope, through a surgical procedure, to discard eyeglasses or contact lenses for distance vision with minimal risk. Its potential value to the military is to be able to correct the increased myopia of highly trained individuals who can no longer meet the uncorrected vision standards for a specific program, and thereby reducing training costs for replacement of personnel. Similarly, the potential to have an infantry or shipboard force that is independent of eyeglass and contact lens correction for better all-weather, all-conditions performance should be recognized. Other potential values are to prevent a reduction in the pool of potential recruits available for any military occupation that will occur with the increasingly widespread use of PRK in civilians and to increase the available pool of individuals for training for specific highly specialized military missions.
With these potential benefits in mind, a certain amount of caution must be exercised. Long term clinical data are not available. There are gaps in the knowledge of what happens in basic corneal wound healing. Refractive outcomes are not currently 100 percent predictable. Only two machines are currently approved by the Food and Drug Administration for clinical use. Modifications of the procedure are still occurring. The exact benefits and problems concerning military tasking are not known.
The panel encourages the U.S. Department of Defense (DoD) to proceed with well-planned studies of PRK within the military. This may be the only way to answer many of the questions concerning military-specific performance. No studies in the civilian population are available that relate specifically to military performance. Neither are such studies likely to be performed without DoD support, as they would have little meaning to most tasks required of nonmilitary populations in everyday life.
In the following sections, the panel summarizes the findings from the literature that it details in the body of the study, and repeats the specific recommendations for the use of PRK in the military and for additional studies to answer some of the questions that are military specific. As assigned in the Terms of Reference (Appendix B), the panel focussed its work on PRK rather than laser-assisted in situ keratomileusis (LASIK). Some information on LASIK is included, but a detailed literature review of LASIK would require another study. Although the four chapters of the study are attributed to individual authors or groups of authors, all members of the panel have reviewed the entire study and reached consensus on its contents.
CORNEAL AND OCULAR PHYSIOLOGY
Findings
Concerning the efficacy of PRK, the wound-healing responses in animal studies from rabbits and primates indicate that the PRK wounds heal predominantly by corneal fibrosis during the first 6 months after surgery. This leads to a complete regression of the refractive effect of PRK in test animals. Based on our knowledge of human healing patterns following radial keratotomy, we would suspect that healing following PRK is considerably delayed and may require up to 3 years to be initiated and perhaps 5 to 10 years or longer to be completed in some patients. While the physiologic effects of delayed wound healing are unknown, delayed healing in Radial Keratotomy has led to biomechanical weakening of the cornea and the development of hyperopic shifts in a subset of patients with a higher number of, and more centrally placed, incisions. Although a similar mechanical effect following PRK is not likely, due to the abrasive rather than incisional nature of the injury, the long term consequences of PRK and its effect on mechanical stability, corneal physiology and refractive stability remain a major concern that needs further study. Current clinical studies reporting results from 6-months and 1-year follow-ups are inadequate and do not assess the full potential of PRK to undergo considerable regression. In addition, clinical studies generally are performed on older individuals who are known to show less aggressive healing responses and undergo less regression. The military population at risk may be considerably younger, (i.e., in their twenties, and they should experience more exuberant healing and greater regression). Also, women of child-bearing age are at greater risk for regression during pregnancy, another concern for the military population.
Concerning the safety of PRK, incomplete regeneration of the epithelial basement membrane in animal studies is a major concern. The long-term effect of an abnormal basement membrane on epithelial and stromal differentiation and function is not known, and the potential risk for infection and epithelial erosion needs to be seriously considered. In addition, corneal wounds that remove the basement membrane heal by stromal fibrosis with deeper injuries, producing greater fibrosis. The presence of corneal fibrosis will clearly effect corneal transparency; the long-term clinical significance is not known. Furthermore, the safety of LASIK surgery is completely unknown.
Concerning predictability, differences in corneal hydration between patients make it unlikely that the accuracy of PRK correction can ever be much better than +1 diopter. Of course that does not take into account the effect of regression that is variable dependent on patient age, sex, and health.
Recommendations
The panel recommends that future studies sponsored by the DoD related to corneal and ocular physiology be directed toward answering the following three questions:
1. What is the cause of regression in patients—epithelial hyperplasia or fibrosis—and how long does it take to stabilize—5 years to 10 years?
2. What is the importance of an abnormal basement membrane? Is there any effect on long-term epithelial and keratocyte differentiation and functions, and what is the risk to later infection?
3. What is corneal haze, how can it be measured objectively and how does it correlate with visual outcomes including glare and low-contrast visual acuity?
Other recommendations for areas meriting further study are as follows (in order of their appearance in the Chapter 2):
1. What is the precise in vivo decrease in corneal thickness after PRK in patients, and how does achieved photoablation correlate with intended correction?
2. What is the risk for the development of early cataract following single and multiple PRK procedures?
3. What is the tolerance of the cornea to shear stresses after LASIK surgery?
4. What is the depth of injury to underlying cells—keratocytes and endothelial cells—in patients, and how does the depth of injury correlate with the development of haze and regression?
5. What is the risk of UV and solar radiation to haze and regression?
6. What is the effect of pregnancy and menopause on PRK?
7. Can growth factors or cytokines alter the post-PRK repair process?
8. What is the early inflammatory response and how does it correlate with haze and regression?
9. What is the mechanism of collagen fibrillogenesis and how does it relate to corneal haze?
10. What is the risk of corneal endothelial damage following LASIK?
11. What is the risk of corneal anesthesia following PRK, and how does reinnervation modulate epithelial healing and long-lasting epithelial defects after PRK? How does corneal anesthesia affect basal tear secretion and subsequent corneal drying and damage?
CLINICAL OPHTHALMOLOGY AND OPTOMETRY
Findings
Although PRK and LASIK are widely accepted and being performed worldwide, many controversies and gaps in knowledge exist with respect to safety, efficacy, and techniques. The following information summarizes the controversies and gaps in knowledge.
PRK versus LASIK
There is much more information on the results of PRK, both in terms of immediate results as well as longer-term follow-up (1–5 years) than LASIK. This gap in knowledge makes discussion of the relative value versus the risks of LASIK incomplete and somewhat speculative. Because of the likelihood of permanent nonattachment of the central corneal flap in LASIK, there may be dangers of flap loss secondary to trauma. LASIK may be superior to PRK for myopia greater than -7.0 diopters, however, and may be associated with more rapid recovery and visual rehabilitation with reduced central scarring. High myopia (greater than
-7.0 diopters) may also be better treated with PRK using second-generation solid-state lasers not currently approved in the United States. Exposure to UV light may have less effects on a patient after LASIK versus PRK.
Most studies have been relatively short term with PRK and limited with LASIK. There is a significant gap in knowledge with respect to potential long-term complications of both of these procedures.
Multizone versus Multiple Pass
There is controversy in the size of the optimal ablation treatment zone in PRK. Best optical results appear to be achieved using a 6-mm zone or larger in PRK in comparison with the previously smaller zones. The ablation zone in LASIK is small by necessity. Spot size appears to be related to "islands" of residual tissue in the ablation zone with large spot size more likely to cause these islands. Multiple pass techniques allow for “drying” and reduce islands. Multizone techniques allow for wider, shallower ablation zones with “smoother” shoulders that may have many positive benefits including reduced haze, reduced halos, better low-frequency contrast acuity, and less regression.
Low Energy versus High Energy
There may be a benefit in using low-energy techniques versus high-energy techniques, especially in higher myopia (less haze, less regression of the treatment effect). Low energy corresponds to solid-state and small-spot-size techniques. These will have more versatility in patient selection (hyperopia and all degrees of astigmatism) and adjustability with potential for computer-controlled, individually designed ablations to create the optimal corneal shape for every patient.
Epithelial Removal
There are several methods available to remove the corneal epithelium prior to PRK: mechanical (FDA approved), laser transepithelial ablation, alcohol debridement. Mechanical debridement of the epithelium with a brush is the most rapid technique but may be less complete than alcohol debridement. The significance of this is not well understood. Laser removal may stimulate less cytokine release from the unaffected edges leading to reduced apoptosis and scarring.
Epithelial Healing
There is little data to support the safety and efficacy of contact lenses in treating epithelial defects. Current practice includes contact lens wear to reduce pain, but is recognized to prolong the epithelial defect for approximately 1 day. In addition, ocular surface diseases such as rosacea, tear dysfunction, and obstructive meibomitis are known to retard healing and should be treated prior to PRK. Use of cytokines to retard apoptosis or to stimulate epithelial growth may prove beneficial.
Contact Lens Wear
Because PRK does not alter the peripheral corneal curvature (unlike RK which flattens it), patients can resume contact lens wear if required. There are no reported problems in the ability to wear contact lenses successfully after PRK, if needed.
Steroid Use After PRK
The use of topically applied corticosteroids following PRK is controversial because of potential complications such as steroid-induced glaucoma, cataracts, and herpes simplex infection. Currently, topical steroids are commonly used in the United States, but not by some European surgeons. More information is necessary to determine the optimum approach.
Spatial Contrast Sensitivity
The loss of low spatial contrast sensitivity following PRK during the first postoperative year is well described in the literature. There are no data that adequately address the issue of whether or not this is relevant to performance or is predictive of decreased visual function.
Performance Standards Pre- and Post Operative
It is critical to know how the performance of an individual is affected by PRK with respect to specific military duties. Pre- and postoperative tests need to be developed to assess the impact on specific areas such as nightdriving, pilot function, rifle performance, and SEAL performance.
Recommendations
The panel recommends that future studies sponsored by the DoD related to clinical ophthalmology and optometry and PRK address the following three research requirements:
1. What is the residual need for glasses or contact lenses following PRK or LASIK? (research requirement 3-1)
2. What is the impact of reduced low-contrast and night vision acuity following PRK or LASIK and how long does it persist? A particularly relevant area for research involves the effects of PRK and LASIK on night driving. Actual driving simulators with a control group may be best to make this assessment because devices that simulate only the visual experience in a narrow view forward (such as the mesoptometer) may be inadequate to assess the entire dynamic visual field. (research requirements 3-5 and 3-10)
3. What is the effect of hypo/hyperbaric environments on LASIK? (research requirement 3-7)
Other recommendations for areas meriting further study are as follows (in order of their appearance in Chapter 3):
1. What are the best measures of visual performance for military tasks (what should the performance standards be for specific tasks for groups such as SEALs, special forces, pilots) since high-contrast Snellen acuity, the typical standard, does not describe the extremes of visual demand? (research requirement 3-2)
2. How long does it take to recover usable vision following PRK or LASIK? (How long would a member be off duty following PRK?) (research requirement 3-3)
3. How long does it take to recover to best spectacle-corrected visual acuity (BSCVA) or best uncorrected visual acuity (UCVA) following PRK or LASIK? (research requirement 3-4)
4. What is the pattern of return to adequate resistance to lateral (shear) forces with healing following LASIK? (Would the flap withstand wind shear as with parachuting or blast forces?) (research requirement 3-6)
5. What has been the positive and negative experience of police and fire-fighting agencies that have authorized PRK or LASIK? (research requirement 3-8)
6. How many patients with BSCVA of 20/20 before PRK or LASIK have 20/25 or worse BSCVA following PRK or LASIK? (research requirement 3-9)
7. How long does the refraction remain stable? Is there a difference in stability between PRK and LASIK? (research requirement 3-11)
8. What should the standard analysis be for determining the effectiveness of astigmatic surgery? (research requirement 3-12)
PRK AND VISUAL FUNCTION
Findings
A summary of the panel’s findings on PRK and visual function are listed below, by category.
Overcorrection and Undercorrection
Virtually all studies of refractive stability following the PRK procedure show a common pattern of post-PRK refractive error. The refractive error immediately following PRK is usually hyperopic (overcorrection), but during the first 2 or 3 months, the manifest refraction gradually becomes less hyperopic (more myopic) again. In some cases, this myopic regression results in emmetropia, but in others, it leaves the patient hyperopic. In most studies, the final Rx after regression has reverted to myopia. In most studies, the refractive error stabilizes after about 3 months, but this finding is not universally observed, and some studies report continuing regression in high myopes over a 1-year period.
From the literature, it appears that success rates are improving: Final post PRK Rx’s are getting closer (on average) to emmetropia.
Because ablation depth will vary with corneal hydration and healing processes, and because both can vary between eyes, it is unlikely that the standard deviations in achieved Rx will be able to be reduced to zero in the same way that the mean Rx can be refined.
Results to date seem to indicate that large initial overcorrections, large myopic regressions, and large residual myopias can be avoided by using a larger ablation zone. We have not found any reasonable explanation for why larger ablation zones should produce improved refractive outcomes.
Stability of the post-PRK refraction has been studied under high-altitude conditions known to produce large hyperopic shifts in post-RK eyes. Unlike post-RK eyes, no hyperopic shifts were observed in the PRK eyes during a 3-day exposure to a hypobaric environment.
Use of large (6-mm) ablation zones seems to correct some serious problems with earlier data collected with 4-mm ablation zones, and post-PRK refractions are stable and close to emmetropia.
Reliance on direct corneal measurement should be utilized to see if the procedure actually changes the cornea by the desired amount because post-PRK Rx is an indirect measure that can be influenced by pupil size.
Changing the Shape of the Cornea
Post-PRK eyes do experience halos when the pupil dilates at low light levels (78 percent in the early postoperative period).
The dioptric step introduced into the peripheral optics makes the eye more myopic at the pupil margin (outside of ablation zone) than at the pupil center (within the ablation zone). This bifocal nature of the post PRK cornea is also likely to be present in a post-LASIK cornea.
Topographical studies report myopic “islands” in the center of the ablation zone, which may be absent with newer lasers. It is not clear what produced them or why they generally disappeared after 3 months.
Corneal Transparency
The ablation process causes the usually very transparent corneal stroma and corneal epithelium to lose some of its transparency. This transparency loss peaks during the first month(s) and declines possibly back to normal levels at 3+ months. The method for removing the epithelium prior to the stromal ablation may also influence the amount of transparency loss. The biological causes need to be clearly identified before a rational approach to eliminating them can be initiated.
The glare effects peak early in the post-operative period (2–3 months, and are virtually absent at 1 year in most eyes). However, there are persistent glare and haze problems that linger in some eyes. Currently there is no clear hypothesis to explain why some eyes have persistent haze.
Night Vision
The change in pupil size and the changes in the visual stimulus can have significant interactions with the optical effects of PRK. Large pupils will include the edge of the ablated zone. The edge of the ablated zone includes (1) increased haze (according to one study) and (2) a huge dioptric step (equal to the step between post-PRK Rx and pre-PRK Rx) that creates a bifocal visual system with decreased image quality and noticeable annular blur rings (halos) around bright light sources. Reduced light levels lead to reduced signal-to-noise ratios and reduced visual contrast sensitivity (and resolution). Thus any additional reductions in contrast sensitivity caused by PRK may have added impact at night where more targets are already close to contrast threshold.
In combining the two effects (pupil dilation and the accompanying reduction in image quality) with the inherently larger impact PRK will have on low-light vision, it is reasonable to expect that the impacts of PRK on vision will be maximal at night. The presence of bright light sources in the nocturnal environment will have an additional detrimental effect on any eye with a scatter source such as the subepithelial losses of transparency present in all eyes during the 1–3-month post-operative period after PRK.
One study reported large reductions in low-contrast acuity in the presence of glare, problems with halos (100 percent post-PRK surgery, and 45 percent at 1 year), increased glare sensitivity that was still present in 66 percent at 1 year, and 19 of 26 patients would have failed the German night driving test based on these results and thus would not be licensed to drive at night in Germany. It is worth noting that studies of night driving point out that high-contrast visual acuity measurements taken at high light levels overestimate visual performance under night driving conditions.
Vision with Imaging Systems
Imaging systems such as image intensification devices (e.g., night vision goggles) pose a special problem for PRK. These devices are designed to multiply the photons from a nighttime scene. In doing so they amplify an inherently noisy signal (reducing the light level is equivalent to reducing the signal-to-noise ratio because of the Poisson nature of photon distributions) and the amplification process adds its own noise. This reduced signal-to-noise ratio is compounded when the object has a low-contrast target (such as a sand dune in the desert) because a low-contrast object also has a low signal to noise ratio. These three factors (low light level, image intensification, and low contrast) already render some objects invisible that we might expect to see because of the high mean intensity of the night vision goggles. Any additional reduction in the signal-to-noise ratio in these marginal conditions could render even more stimuli invisible. It is therefore possible that a PRK procedure that allows high-contrast objects viewed in daylight to be seen with ease may render many low-contrast objects invisible when viewed with night vision goggles.
Recommendations
The panel’s most important recommendations related to PRK and visual function are summarized below.
1. The most daunting, yet most critical task in assessing the consequences of PRK for military operations is the design and implementation of appropriate performance tests. We note that many critical military operations are conducted under extreme conditions in which the visual system's ability to pick up the required information is pressed to a variety of limits. We recommend that performance tests be designed and conducted that require subjects to perform representative tasks under visual conditions that mimic those in the specific military situations concerned. It would be best to conduct such tests using within-subject designs so that the same subjects can be tested in the same tasks with different simulated optical degradations. The more important results should ultimately be confirmed with PRK subjects and controls.
2. Given the military needs for optimal night performance, and the paucity of data (one abstract from the Association for Research in Vision and Opthamology) on night vision with PRK, an obvious gap in the experimental literature exists. There is a desperate need for good psychophysical data and controlled experiments on the effects of PRK on night vision with and without amplification.
3. We strongly recommend that any future PRK studies evaluate visual function using measures other than high-contrast letter acuity when assessing the suitability of PRK for the military and that those measures be tailored to the specific military task in question. Some aspects of the viewing situation are particularly important to measure and examine in the research effort. These include the luminance, contrast, and spatial frequency content of the targets; the subject's pupil diameter; and the position and intensity of glare sources.
4. Considering the inconsistencies in the literature and the inadequate level of optical data on PRK, we recommend a single study that monitors (1) Rx, (2) corneal thickness changes, and (3) corneal curvature changes. Using a simple model, these data can confirm the success or failure of PRK in achieving what it is supposed to achieve. We can find no study that has formally tested the refractive effects of PRK in this way.
5. Future studies should employ better optical methods (compared with inferred optical changes based on indirect measures used with normal eyes), including proper assessment of contrast sensitivity before and after PRK. We suggest that the military consider simulating the optical degradations associated with PRK as another possible method for evaluating the impact of PRK on visual function.