PTK can be a highly effective therapy for a central corneal pathologic process arising beneath a relatively smooth epithelial surface, as in many anterior stromal dystrophies (31). When the anterior corneal surface is smooth, or the lesion in Bowman’s membrane or anterior stroma is not causing significant surface irregularity, the goal of PTK is twofold. The first aim is to translate this smooth surface to a level deeper than the corneal lesion, and the second is to minimize the induced refractive error (6). In these cases, epithelial removal is best accomplished by the laser rather than by mechanical debridement. The epithelium itself acts as a masking agent to allow irregularities in Bowman’s layer and the superficial stroma to be differentially ablated. Some surgeons prefer to use a masking fluid on intact epithelium, ablating until the blue fluorescence disappears, signifying the point at which all epithelium has been ablated and stromal ablation begins (32,33). Manual epithelial debridement has been shown to be problematic for corneal dystrophies such as Reis-Buckler. Lawless et al. noted that manual epithelial removal in an eye with Reis-Buckler dystrophy was often patchy and the end point for complete removal difficult to assess (34, 35, 36).

Because corneal dystrophies usually manifest as a diffuse pathologic area, a large spot size, measuring 6 to 7 mm, centered on the entrance pupil, can be used to ablate both epithelium and underlying anterior stroma. Initial depth of ablation should be 75% of the estimated depth of the corneal lesion. After the initial ablation, the surgeon examines the patient at the slit lamp to evaluate the effect, and to assess if further ablation should be done. The “ablate and check” method is performed until the desired bulk of diseased cornea is removed (31). The depth of the opacity can be difficult to assess with this technique, and care should be taken not to exceed the planned preoperative depth of ablation as determined by careful slit-lamp examination and optical pachymetry.

The surgeon must remember that as ablation depth increases, the change in refractive error also increases, and care should be taken to ablate the minimal amount of lesion needed to alleviate patient symptoms (6,37). Undertreatment is preferable to overtreatment because the procedure can be repeated.

Mechanical debridement is the preferred method for epithelial removal when treating recurrent epithelial erosions or any disorder in which the epithelium is deemed irregular compared with the underlying Bowman’s membrane. This can be accomplished manually with a Bard-Parker blade (38). The epithelium is removed over the area of erosion. If the visual axis is involved, it may be wise to include this area so as not to induce irregularities into the ablation profile. Masking fluid, such as 1% hydroxymethylcellulose, 0.5% tetracaine, or Tears Naturale, may be applied after manual debridement to smooth the surface before ablation (6). The ablation zone is chosen to encompass the area of debrided epithelium. A depth of 3 to 10 μm is usually sufficient to remove a portion of Bowman’s layer (6,31,37).

To smooth the resultant stromal bed for epithelialization and to reduce the magnitude of induced hyperopia after performing a large central ablation, a transition zone can be created. Approximately 1 D of hyperopic shift is induced for every 20 μm of stromal ablation (6). Sher et al. used a circular motion of the eye to perform a “smoothing technique” (34). The Summit excimer laser trials used a related “polish technique” whereby the patient’s head was moved in a brisk circular manner under the laser (39,40). Subsequently, Stark et al. described a “modified taper technique” whereby the surgeon moves the eye in a circular fashion under the laser, treating the periphery of the ablation zone with a 20-μm deep, 2-mm diameter spot size. The additional peripheral ablation limits induced hyperopia by approximately 2 to 4 D (7,40). Alternatively, a hyperopic photorefractive keratectomy (PRK) ablation may be used after PTK to minimize induced hyperopia. In this circumstance, epithelial ablation should be carried out to 8 mm. Combining PTK with PRK is technically easier and potentially more precise than the polish techniques; however, it requires additional costs for the hyperopic card and removes additional central as well as peripheral tissue. From previous studies, it is known that approximately 1 D of hyperopic shift is induced for every 20 μm of stromal ablation (6), and the depth of both the PTK and PRK ablations should be considered in the total tissue removed. Care should be taken to avoid too much thinning of the peripheral cornea because a corneal transplantation may be considered at a later time. PTK combined with PRK can also be considered for PTK candidates who have concurrent myopia.

Allowing the precise removal of a predefined amount of corneal tissue, PTK is well suited for ablation of elevated scars and nodules. Such lesions can be seen in keratoconus, Salzmann’s degeneration, and posttraumatic or infectious scars. Decreased vision may be secondary to opacification, surface irregularities, or associated refractive errors (6). Again, disease limited to the anterior stroma is most amenable to PTK. Unlike ablation of smooth corneal lesions, with the treatment of elevated nodules or scars, the goal is to smooth the lesion selectively, while simultaneously limiting refractive change. It is best to measure the densest portion of the scar in the optical zone using the optical pachymeter, and to limit ablation to the depth determined before surgery. Because the excimer laser removes surface tissue and does not automatically smooth contour irregularities, any preexisting irregularities will be duplicated at a deeper level if ablation rate and spot size stay constant (41). In addition, scars or calcific lesions may ablate more slowly than normal corneal tissue, and the excimer laser can induce respective peaks and valleys in the contour that were not present before treatment (42, 43, 44). Finally, selective ablation of central lesions can lead to large hyperopic shifts, whereas ablation of peripheral lesions can induce myopia (6).

The importance of surface modulators in these circumstances to aid in smoothing the contour of the areas to be treated cannot be overestimated. The smooth contour created by these modulators can be translated to the new postablation surface (6). Fluid masking agents, which fill valleys in the corneal surface, are probably the most commonly used type of surface modulator. These fluids include 0.9% saline (Unisol), 1% carboxymethylcellulose sodium (Celluvisc), and 0.3% hydroxypropylmethylcellulose 2910 with 0.01% dextran 70 (Tears Naturale II). These solutions all exhibit sufficient absorbance by the 193-nm ArF laser, although 0.3% hydroxypropylmethylcellulose 2910 and a 0.1% dextran 70 solution showed greater absorbance than 0.9% saline (41). The viscosity of a masking fluid should be high enough to remain in surface depressions without run-off and low enough to coat the surface uniformly. A thin layer of fluid should be applied with a damp cellulose sponge. A “sight and sound” method can be used to assess the adequacy of the application. Because the excimer laser turns many masking fluids white, a white area overlying a peak indicates that too thick a layer of fluid has been applied. Because ablation of the fluid emits a soft “click” versus the loud “snap” of naked cornea, a louder treatment than predicted signifies deficient coverage of fluid on the cornea (39). The fluid may need to be reapplied several times to attain the desired layering throughout the procedure.