Even though the advent of CXL in the 1990s established it as an effective means of keratoconus stabilization, the evidence overall reveals minimal topographic regression and little visual improvement in keratoconus after treatment. This has prompted several practitioners, ourselves included, to investigate means to retain the disease-stabilizing aspects of the CXL treatment and to obtain significant improvement in corneal irregularity and, hence, vision.
Several pioneering studies have evaluated visual rehabilitation in stable forme fruste keratoconus utilizing PRK alone, prior to the development of CXL.1-3 Despite initial fears, there was no evidence of progressive ectasia after PRK, and this laid the foundation for developing a synergistic CXL-plus-PRK therapy for more advanced progressive keratoconus.4
It is important to distinguish among three distinct applications of combinations of CXL and PRK, all with differing therapeutic goals (Table 1); the discussion below is limited primarily to frank keratoconus. We have previously described our approach, which is to combine nontopographic transepithelial PRK without mitomycin C plus immediate CXL.5 While approaches vary, common to all is the finding of greater visual improvement than would be expected from CXL alone, as confirmed by recent comparative studies of PRK plus CXL versus CXL alone.5
ISSUES SURROUNDING PRK PLUS CXL
The evolution of various approaches to combining these two therapies raises several issues.
Kanellopoulos and Binder initially proposed CXL followed by delayed topographic PRK.4 Theoretically, this approach avoids superimposing the unpredictable CXL effect on the topographic correction. However, our experience,6 in keeping with subsequent work by the groups of Kanellopoulos7 and Kymionis,8,9 suggests that there is an additive effect when the two procedures are performed simultaneously. Thus, comparing sequential (CXL first and PRK later) versus simultaneous (PRK with immediate CXL), Kanellopoulos found superior results in UCVA, BCVA, keratometry, and haze with the simultaneous approach.7
There are some clues to why this additive effect occurs when the procedures are performed simultaneously. PRK at an interval after CXL ablates the maximally crosslinked anterior cornea, and the ablation rate is also reduced, altering correction. The most important effect, in our opinion, relates to the potential enhanced crosslinking effect when CXL is applied immediately after ablation, as discussed below.
Our technique of nontopographic transepithelial PRK and immediate CXL is based on the clinical premise that the topographic effect is much greater than would be expected as the sum of the effects of a minimal ablation and CXL. A clinical example of our technique (Figure 1) illustrates far greater change than would be expected from adding the effect of CXL to a stromal ablation of less than 50 µm. This potentiation of CXL after surface ablation has been demonstrated by Kymionis et al in an approach termed the .7 The authors showed an enhanced effect when PTK, rather than mechanical epithelium removal, was performed before CXL. Moreover, Kymionis et al also demonstrated a deeper demarcation line under the ablated area, suggesting enhanced CXL effect.9
It is our opinion that topographic planning based on the cornea’s preoperative state before CXL is fraught with sources of unpredictability, particularly the effect of CXL itself. This is compounded by poor correlation between topographic and refractive findings and by the limited topographic correction obtainable when a limited ablation is performed. Our technique, thus, foregoes profile-based correction of higher-order aberrations. Notably, the visual acuity gains in our series do not vary greatly from those in series using topographic correction.6-8,10
In several of the initial series cited above, alcohol or mechanical epithelial removal were used. However, like other more recent descriptions, our protocol incorporates a transepithelial treatment in order to obtain a topographic corrective effect based on the differential epithelial profile around the cone, which has been shown to have a so-called donut distribution; thus, the masking effect of the epithelium would tend to regularize the cone.
Although mitomycin C has been used in keratoconic PRK, we theorized that CXL-induced keratocyte depopulation may offset a need for chemotherapy, particularly as the transepithelial treatment causes less haze. Our data and those of a similar study by Stojanovic both showed insignificant haze.10
CONCLUSION
Approaches combining PRK or PTK and CXL for the management of frank keratoconus show encouraging results. We expect that further study and refinement of treatment parameters will yield further benefits. For the present, PRK remains the most well-established corneal refractive procedure and the only one with an established safety profile allowing its use in ectatic disorders of the cornea.
Ioannis Aslanides, MD, PhD
• Medical Director, Emmetropia Mediterranean Eye Institute, Crete, Greece
• i.aslanides@emmetropia.gr
• Financial disclosure: None
Hatch Mukherjee, FRCOphth
• Consultant Ophthalmologist, Essex County Hospital, Colchester, United Kingdom
• hatch.mukherjee@gmail.com
• Financial disclosure: None
