Chapter 23 – Laser Subepithelial Keratomileusis (LASEK)
DIMITRI T. AZAR
ROBERT T. ANG
• Laser subepithelial keratomileusis (LASEK) is a surgical technique that combines features of photorefractive keratectomy (PRK) and laser in situ keratomileusis (LASIK) in which a corneal epithelial flap is elevated and replaced at the end of the procedure.
• Following preplaced epithelial marking, dilute ethanol is applied and a hinged epithelial flap is created by peeling the loosened epithelium as a sheet. After laser ablation, the flap is repositioned over the ablated stroma.
• Although LASEK may not consistently provide immediate and complete epithelial coverage, the early results seem promising, especially for the potential applications in wave front–guided keratorefractive surgery. 
In laser subepithelial keratomileusis (LASEK), loosening and replacing the corneal epithelium can be performed using several techniques.       Photorefractive keratectomy (PRK) and laser in situ keratomileusis (LASIK) are currently the most popular procedures in refractive surgery. PRK is a relatively safe procedure; its major limitations are postoperative pain, subepithelial haze, and prolonged visual rehabilitation. The epithelial removal done prior to laser ablation in PRK is believed to be the major factor contributing to these drawbacks.  LASIK offers more comfort, faster visual rehabilitation, and minimal haze, but it has its own set of complications, predominantly related to the flap. These include free caps, incomplete pass of the microkeratome, flap wrinkles, epithelial ingrowth, flap melt, interface debris, and diffuse lamellar keratitis.  
In selected patients for whom PRK would be the recommended procedure, such as patients with thin corneas, patients with lifestyles or professions that predispose to flap trauma including athletes in contact sports and military personnel, and patients with low myopia who are at a lower risk for subepithelial haze, laser epithelial keratomileusis (LASEK) may be a viable alternative.  
LASEK theoretically offers the advantage of avoiding the flap complications of LASIK and also addresses the drawbacks of discomfort and delayed recovery associated with conventional PRK. Epithelial sheet viability and adhesion are the basis for achieving the potential advantages of LASEK.         The indications and contraindications for LASEK are summarized in Box 23-1 .
In patients with corneal pachymetry of 500?µm or lower and in patients with asymmetric corneal curvature, the use of the Azar-Lu MEEI keratoconus classification ( Fig. 23-1 ) is helpful in
LASEK Indications and Contraindications
Thin corneal pachymetry
Epithelial irregularities (M/D/F changes)
LASIK complications in the contralateral eye
Predisposition to trauma
Irregular astigmatism (corneal topographical abnormalities not qualifying as keratoconus [ Fig. 23-1 ])
Patient concern about postoperative pain
Hyperopia and hyperopic astigmatism
From Azar DT, Ang RT. Laser subepithelial keratomileusis: evolution of alcohol-assisted flap surface ablation. Int Ophthalmol Clin. 2002;42:89–97.
identifying patients with keratoconus in whom surgery should be avoided.
Patients undergo routine preoperative evaluation in a manner similar to that for other refractive surgical procedures. These include uncorrected visual acuity (UCVA), best-corrected visual acuity (BCVA), manifest and cycloplegic refraction, ocular dominance, keratometry, tonometry, pachymetry, slit-lamp examination, aberrometry, and computerized videokeratography.   
Several surgical techniques have been described including minor modifications of our original technique, which is illustrated in Figure 23-2 . Camellin and Cimberle and Vinciguerra and Epstein have described similar techniques with encouraging results.
Thirty minutes before surgery, the eye receives trimethoprim sulfate 1?mg/ml, polymyxin B 10,000?U/ml (Polytrim; Burroughs Wellcome, Research Triangle, NC) or ciprofloxacin (Ciloxan; Alcon Laboratories, Fort Worth, TX). 
A sterile drape is placed around the eye. Then one drop each of topical 0.5% proparacaine (Ophthetic; Allergan, Irvine, CA) and 4% tetracaine (formulated in the MEEI pharmacy) are instilled and a lid speculum applied. The cornea is marked with overlapping 3?mm circles around the corneal periphery, simulating a floral pattern. In our early cases a 7?mm optical zone marker (model E9011 3.0; Storz, St. Louis, MO) was used to delineate the area centered around the pupil. Gentle pressure was applied on the cornea while the barrel of the optical zone
Figure 23-1 The Azar-Lu Keratoconus Classification. (From Melki SA, Azar DT. LASIK complications: etiology, management, and prevention. Surv Ophthalmol. 2001;46:95–116.)
marker was filled with two drops of 18% ethanol (dehydrated alcohol, 1?ml ampules; American Reagent Laboratories, Shirley, NY). In more recent cases an alcohol dispenser consisting of a customized 7?mm semisharp marker (ASICO, Westmont, IL) attached to a hollow metal handle served as a reservoir for the 18% alcohol. Firm pressure is exerted on the central cornea and a button is pushed on the side of the handle, releasing the alcohol into the well of the marker. After 25 seconds, the ethanol is absorbed using a dry cellulose sponge (Weck Cell or Merocel; Xomed, Jacksonville, FL) to prevent alcohol spillage onto the epithelium outside the marker barrel. 
Modified Vannas scissors are inserted under the epithelium and traced around the delineated margin of the epithelium, leaving 2–3 clock hours of intact margin, preferably at the 12 o’clock position. The loosened epithelium is peeled as a single sheet using a blunt spatula or a Merocel sponge, leaving a flap of epithelium with the hinge still attached. The laser ablation is initiated immediately thereafter using an excimer laser. After ablation, a 30-gauge anterior chamber cannula is used to hydrate the stroma and epithelial flap with balanced salt solution. The epithelial flap is replaced on the stroma using the straight part of the cannula under intermittent irrigation. Care is taken to realign the epithelium flap using the previous marks and to avoid epithelial defects. The flap is then allowed to dry.
Topical medications are then applied: diclofenac sodium 0.1% (Voltaren Ophthalmic; Ciba Vision Ophthalmics, Duluth, GA) and tobramycin 0.3% dexamethasone 0.1% ointment (TobraDex; Alcon Laboratories) or diclofenac sodium 0.1% (Voltaren Ophthalmic; Ciba Vision Ophthalmics), ciprofloxacin (Ciloxan; Alcon Laboratories), and fluorometholone 0.1% (FML; Allergan America, Hormigueros, PR) drops. A bandage contact lens (Soflens 66; Bausch & Lomb, Rochester, NY) is then placed.
After the procedure, all patients are prescribed oral analgesics and instructed to take them only if needed. The postoperative regimen consisted of tobramycin-dexamethasone ointment (TobraDex; Alcon Laboratories) four times a day for 1 week and prednisolone acetate 1% (Pred Forte 1%; Allergan, Irvine, CA) four times a day for 2 weeks. Artificial tears are also prescribed (on an as-needed basis). The bandage contact lens is removed after complete reepithelialization (at postoperative days 3 and 4). Removal of the contact lens in the first postoperative day risks peeling the epithelial flap with the contact lens.
After exposure to 18% ethanol for 25 seconds, the epithelium was successfully peeled as a single sheet leaving a 2–3 clock hour hinge. The wound edge was sharply demarcated, and no residual islands of epithelium were grossly visible under the microscope. After laser ablation, the folded epithelium was repositioned and the flap edge closely reapproximated to the wound edge. No epithelial defects were visible in all cases prior to placement of the bandage contact lens.  
One day after surgery, an epithelial defect was documented in 30 of the 55 (54.5%) cases that were followed up. On the third postoperative day, 25 of the 30 defects (83.3%) had completely healed, 2 cases still had a defect, and 3 cases did not return for postoperative check. Twenty-five eyes (45.4%) had no defect on the first postoperative day. On day 3, a defect developed in 5 of these 25 (20%) eyes. Three resolved by 1 week; the other two
Figure 23-2 Our LASEK technique. A, Multiple marks are applied around the corneal periphery, simulating a floral pattern. B, An alcohol dispenser consisting of a customized 7 or 9?mm semisharp marker attached to a hollow metal handle serves as a reservoir for 18% alcohol. Firm pressure is exerted on the cornea and alcohol is released into the well of the marker. C, After 25–30 seconds, the ethanol is absorbed using a dry cellulose sponge. D and E, One arm of a modified Vannas scissors (Azar scissors; note the knob at the tip of the lower arm) is then inserted under the epithelium and traced around the delineated margin of the epithelium, leaving a hinge of 2–3 clock hours of intact margin, preferably at the 12 o’clock position. F, The loosened epithelium is peeled as a single sheet using a Merocel sponge or using the edge of a jeweler’s forceps, leaving it attached at its hinge. G, Laser ablation is performed. H, An anterior chamber cannula is used to hydrate the stroma and epithelial flap with balanced salt solution. The epithelial flap is replaced on the stroma using the cannula under intermittent irrigation. I, Care is taken to realign the epithelial flap using the previous marks and to avoid epithelial defects. The flap is then allowed to dry for 2–5 minutes. Topical steroids and antibiotic medications are then applied. J, A bandage contact lens is then placed. (From Feit R, Taneri S, Azar DT, et al. LASEK results. Ophthalmol Clin North Am. 2003;16:127–35.)
cases were not followed up. After the first postoperative week, all previous defects had healed and no reports of recurrent erosions were noted on subsequent visits.  
One day after LASEK, 32 of 55 (58.1%) patients complained of postoperative pain or took oral medications to relieve pain or discomfort.  At the 3-day follow-up, 24 of the 32 (75%) patients reported pain disappearance without medications, 6 still had pain, and 2 cases were not followed up. Of the 23 (41.8%) cases who did have pain and did not take medications on the first postoperative day, 6 patients reported pain on day 3. One week after LASEK, no further pain episodes were reported. 
On day 1, 25.4% had UCVA of 20/40 or better. This increased to 62.7% at day 3, reaching 100% by 1 week and 1 month. At 3 days 23.5% of patients achieved UCVA of 20/25 or better, improving to 80.4% at 1 month. A postoperative spherical equivalent of 0.50D was achieved in 75.6% of cases by 1 month.
A healing line in the epithelium was observed in 22 of 51 (43.1%) eyes at day 3. The line was still seen in three eyes (16.6%) at 1 week and totally disappeared at 1 month. Subepithelial punctate keratopathies (SPKs) appeared in eight subjects (15.6%) at day 3. Among these eight cases, one eye still had the SPK at 6-month follow-up. Eight additional cases of SPK were diagnosed from 1 week to 6 months postoperatively. Trace corneal haze was observed as early as 3 days in one eye but disappeared at 3 months. At 1 month, trace to slight corneal haze was noted in nine eyes. Of the nine, five cleared by 3–12 months postoperatively whereas two had persistent haze at 6 months. Five new cases of haze were reported at 3 months, four of which resolved after 6 months. 
Transmission Electron Microscopy
We examined several corneal epithelium specimens that were obtained from patients who underwent alcohol-assisted epithelial
Figure 23-3 A, Low-magnification transmission electron micrograph of specimen. The epithelial cells were associated tightly with flattened superficial cell layers. B–D, Higher magnification transmission electron micrographs of the basal cells and basement membrane region. B, The irregular structure of the basement membrane was observed under the basal cell layer. C, The basal cells were attached to each other through desmosomes (white arrowheads). D, The ultrastructure of the basement membrane was irregular (black arrowheads), but hemidesmosomes were abundant and intact (white arrowheads). (Bar, A: 10?µm, B, C, D: 1?µm.) (Reproduced from Chen CC, Chang JH, Lee JB, et al. Human corneal epithelial cell viability morphology after dilute alcohol exposure. Invest Ophthalmol Vis Sci. 2002;43(8):2593–602.)
removal prior to PRK. The specimens were fixed in glutaraldehyde and processed in 0.1?M cacodylate buffer (pH 7.5). The samples were then postfixed in osmium tetroxide, dehydrated in graded alcohols, embedded in Epon 812, and oven-dried at 60°C for 48 hours. Sections 1?m thick were stained with toluidine blue for orientation. Subsequently, ultrathin sections were obtained using a transmission electron microscope (Philips 410; Lico, Bedford, MA).
Electron microscopy showed variability in the histological appearance of the basement membrane zone. Most specimens showed intact epithelial cell layers ( Fig. 23-3 ). Edematous cells and abnormal vacuoles were observed in other specimens. Higher magnification revealed variable configurations of the epithelial basement membranes including normal areas and areas of discontinuities and irregularities in the basement membrane. Basement membrane fragments were still attached to the epithelial basal cells in most specimens. Bowman’s layer and corneal stroma were absent, indicating that the epithelial sheets separate from Bowman’s layer with variable amounts of basal laminae attached to the basal epithelial cell layer. The ultrastructures of desmosomes and hemidesmosomes were normal in most specimens.
The main indications for LASEK are thin corneas and professions or lifestyles such as contact sports and the military that predispose to flap trauma.          Patients with low myopia who are at a lower risk for subepithelial haze may also benefit from LASEK. It is a simple, inexpensive procedure that involves the creation of an epithelial flap after exposure to 18% alcohol for 25 seconds and subsequent replacement of the flap after laser ablation.   It offers the potential advantage of avoiding LASIK-related flap complications and decreasing the epithelial healing time and postoperative pain associated with PRK.
Our LASEK technique evolved from PRK after alcohol-assisted epithelial removal. Abad et al.  showed that alcohol-assisted epithelial removal was a simple and safe alternative to mechanical epithelial removal before PRK. Applying 25% ethanol for 3 minutes, Stein et al. were able to grasp, lift, pull apart, and split the corneal epithelium using two McPherson forceps. Similarly, Shah et al. exposed the epithelium using a dry sponge. These early reports revealed that epithelial removal using 18–25% alcohol for 20–25 seconds was fast, easy, and safe to perform compared with mechanical débridement; that this
concentration can produce sharp wound edges and a clean, smooth Bowman’s layer; and that the central epithelium can be translocated in part or in toto. 
To study the effect of alcohol exposure and mechanical manipulation on corneal epithelium, we carried out electron microscopic studies of specimens obtained after conventional alcohol-assisted PRK. The images revealed that the epithelial cell layer is intact and the epithelial cells are still viable immediately after exposure to alcohol and surgical peeling. The presence of the basement membrane attached to the basal epithelial cell layer indicates that the point of separation was likely to be within the basement membrane or between the basement membrane and Bowman’s layer. In addition, the adherence of the basement membrane to the basal layer of the epithelium is significant because it is believed that the basement membrane provides the stability and support that keep the epithelium intact even after manipulation. The preservation of the hemidesmosomes in the basal epithelial layer provides a structure that may promote the adhesion of a viable epithelium to the ablated stroma.
The Camellin LASEK technique differs slightly from our technique: a preincision is made using a Janach trephine (Como, Italy) and specialized microhoe, 20% alcohol solution is instilled using a small silicone irrigator, the corneal surface is dried after 30 seconds, and then the epithelium is detached using the short side of a hockey spatula and returned after laser ablation. From our experience, we have learned that the technique may not require specialized instruments but requires several key steps for consistent epithelial flap creation and replacement. Pretreatment with 4% tetracaine prior to alcohol exposure was helpful in loosening the epithelium and lessening intraoperative discomfort. Placement of overlapping corneal marks was crucial in ensuring correct epithelial alignment and avoiding irregular epithelial placement and mismatch. We used an alcohol dispenser, but any optical zone marker with a barrel could be used to expose the epithelium to alcohol and avoid spillage. A jeweler’s forceps to delineate the wound edge and locate the dissection plane and a dry, nonfragmenting, cellulose sponge to peel the epithelial sheet are easily available instruments that can be used to create the flap. The flap can be repositioned with an irrigating cannula under intermittent hydration using the preplaced corneal marks as a guide. No overlap of the flap and wound edge was observed, which would have been attributable to stretching of the flap during peeling or overexpansion due to generous hydration. A 5-minute waiting period was adequate to allow adhesion of the epithelial flap to the stroma. Although LASEK required approximately 10 additional minutes of surgical time, including 25 seconds for alcohol exposure, 1 minute for peeling, 2 minutes for flap repositioning, and 5 minutes drying time, it was easily performed and reproducible. The epithelial flap and hinge remained intact despite handling and manipulation with the irrigating cannula.
After treating 249 patients, Camellin and Cimberle observed that intraoperative flap management was easy in 60% of cases, average in 28%, and difficult in 12%. No pain was experienced by 44% of their cases in the first 24 hours after surgery, and 80% of preoperative BCVA was achieved by 90% of their patients 10 days postoperatively. Vinciguerra and Epstein reported only postoperative grittiness with no pain episodes in 432 eyes treated; 89% achieved refractive stability at 1–2 weeks. The 12-month mean spherical equivalent was -0.10 ± 0.7D. Haze was not found to exceed trace.
Using the techniques we have described,   sloughed epithelium was observed under the bandage contact lens in most patients with epithelial defect at day 1. Visual stability of 20/40 UCVA was consistently achieved after 1 week. This was consistent with the results of Camellin and Vinciguerra. SPK was observed in some patients, indicative of postrefractive surgery dry eye. The use of artificial tears was helpful in relieving dry eye discomfort. Trace to slight haze was noted in some patients but was found to disappear after several months. This may be due to the low myopic corrections in our study population, making it attractive to patients with low myopia who may want to avoid the risk of flap complications. 
The main limitation of LASEK reports is the very small study populations because of the relatively limited indications for this procedure. We have persistently offered it to our patients with low myopia, but most still prefer LASIK, with which the absence of postoperative pain and epithelial defect and fast visual rehabilitation can be more consistently achieved. Two patients initially had LASEK performed 6 years ago. One patient was lost to follow-up after 1 year. The other patient showed long-term stability of refractive outcomes 5 years after surgery.
The main disadvantages of this procedure remain the unpredictable postoperative pain and epithelial healing. Even after ensuring that no epithelial defects were present at the end of each procedure, deepithelialized areas were still observed in more than half of the cases 1 day after surgery. This was accompanied by a similar number of reports of postoperative pain. Because pain is the most compelling drawback of PRK, rapid epithelial coverage is paramount to ensure the patient’s comfort in the immediate postoperative period. LASEK cannot guarantee that it can achieve this consistently.  
1. Azar DT, Ang RT, Lee JB, et al. Laser subepithelial keratomileusis: electron microscopy and visual outcomes of flap PRK. Curr Opin Ophthalmol. 2001;12: 323–8.
2. Azar DT, Ang RT. Laser subepithelial keratomileusis: evolution of alcohol assisted flap surface ablation. Int Ophthalmol Clin. 2002;42:89–97.
3. Camellin M, Cimberle M. LASEK technique promising after 1 year of experience. Ocul Surg News. 2000;18(14):1, 14–17.
4. Vinciguerra P, Epstein D. Laser epithelial keratomileusis (LASEK): one-year results of a new excimer refractive procedure. Presented in the 104th AAO Meeting; Dallas, TX; October 2000.
5. Abad JC, An B, Power WJ, et al. A prospective evaluation of alcohol-assisted versus mechanical epithelial removal before photorefractive keratectomy. Ophthalmology. 1997;104:1566–75.
6. Abad JC, Talamo JH, Vidaurri-Leal J, et al. Dilute ethanol versus mechanical debridement before photorefractive keratectomy. J Cataract Refract Surg. 1996;22: 1427–33.
7. Stein HA, Stein RM, Price C, Salim GA. Alcohol removal of the epithelium for excimer laser ablation: outcomes analysis. J Cataract Refract Surg. 1997;23:1160–3.
8. Shah S, Doyle SJ, Chatterjee A, et al. Comparison of 18% ethanol and mechanical debridement for epithelial removal before photorefractive keratectomy. J Refract Surg. 1998;14:S212–14.
9. Feit R, Taneri S, Chen CC, et al. LASEK techniques and outcomes. Ophthalmol Clin North Am. In press.