A Randomized Prospective Clinical Trial Comparing Laser SubepithelialKeratomileusis and Photorefractive Keratectomy

Objective  To compare laser subepithelial keratomileusis (LASEK) and photorefractivekeratectomy (PRK) in different eyes of the same subjects for subjective painlevel, visual acuity, and corneal epithelial healing.

Design  Prospective, randomized, double-masked study.

Setting  David Grant US Air Force Medical Center, Travis Air Force Base, Calif.

Participants  A convenience sample of 30 active-duty military members with mild tomoderate myopia.

Methods  All patients had LASEK performed in one eye and PRK performed in thecontralateral eye; the order of surgical procedures (ie, right eye first orleft eye first) and the choice of procedures (ie, PRK in the right eye andLASEK in the left eye or LASEK in the right eye and PRK in the left eye) weredetermined in advance using a block randomization table.

Main Outcome Measures  The primary outcome measures were subjective pain level and the rateof corneal epithelial defect recovery. Postoperatively, subjects were evaluatedfor their subjective pain level, visual acuity, and corneal healing (ie, epithelialdefect size) during the first week and up to 30 days after undergoing thesurgical procedures.

Results  There were no significant differences in subjective pain levels betweenthe LASEK- and PRK-treated eyes on postoperative days 1, 2, or 3 (P>.05) or in visual acuity on postoperative days 3, 7, or 30 (P>.05). There was a statistically significant (P<.001) smaller median epithelial defect in the LASEK-treated group(1.0 mm²) compared with the PRK-treated group (16.0 mm²)on postoperative day 1. However, by postoperative day 3, the PRK-treated group(0.0 mm²) showed significantly (P<.001)smaller epithelial defects compared with the LASEK-treated group (4.0 mm²). By postoperative day 7, epithelial defects were undetectable inany subjects in either group.

Conclusions  Laser subepithelial keratomileusis and PRK have similar postoperativepain thresholds and visual acuity recordings. However, the epithelial healingpattern for LASEK and PRK differs. No additional clinical benefit is seenfrom the LASEK procedure relative to the PRK procedure.

Laser subepithelial keratomileusis (LASEK) is a newly advocated laserrefractive procedure for patients with myopia in whom photorefractive keratectomy(PRK) or laser in situ keratomileusis may not be ideal because of their drawbacks.^1,2 These drawbacks may include subepithelialhaze, postoperative discomfort, and prolonged visual recovery for patientswho undergo PRK and who have flap complications and/or diffuse lamellar keratitis,and epithelial ingrowth for patients who undergo laser in situ keratomileusis.^3-6 Laserin situ keratomileusis and PRK also may be unsuitable refractive choices forcertain patients who have corneal thinning or for patients who have higherpotential risk for corneal injuries.⁷ In thesegroups of patients, LASEK has been proposed to be a superior alternative refractiveprocedure.^7-9

In LASEK, a corneal epithelial flap is created using diluted alcohol(18%-20%) placed on the cornea for 30 to 40 seconds.¹⁰ Thereafter,the same excimer laser used typically for PRK or laser in situ keratomileusisis used to ablate the corneal stroma.¹ Theepithelial flap is reflected back on the cornea at the end of the LASEK procedure.Recent studies^2,5,9,11,12 haveshown a reduction in pain, faster visual rehabilitation, better visual acuity,and decreased corneal haze in LASEK-treated patients compared with PRK-treatedpatients; thus, LASEK has been proposed to be a superior refractive procedurecompared with PRK. Another study¹⁰ has alsoshown the vitality of corneal epithelial cells after exposure to the dilutedalcohol. Photorefractive keratectomy, on the other hand, has been well establishedfor the past decade to be the standard traditional method of treating myopia,hypermetropia, and certain types of astigmatism.¹³ Inthe past 2 years, several publications have used comparative control samplesin their study design of the LASEK procedure^14-16;however, these studies either lacked randomization, double masking, or werenot prospective.

In this study, we attempt to answer the question of whether LASEK istruly a better refractive option compared with PRK by comparing these proceduresperformed consecutively in the different eyes of the same subjects. In thisstudy, specific variables such as pain assessment, visual acuity, and cornealepithelial healing recovery period were measured. To date, we are unawareof any randomized double-masked study in the literature comparing the subjectivepain level, visual acuity, and the corneal epithelial healing period of LASEKand PRK consecutively performed in different eyes of the same patients atthe same treatment sitting.

This was a prospective, randomized, double-masked, comparative clinicaltrial performed in the Refractive Laser Center, David Grant US Air Force Center,Travis Air Force Base, Calif. Subjects signed an informed consent documentto undergo a clinical trial comparing LASEK with PRK procedures, which wasreviewed and approved by the facility institutional review board. All surgicalprocedures conformed to the tenets of the Declaration of Helsinki. The voluntary,fully informed consent of the subjects used in this research was obtainedas required by 32 CFR §219 and AFI 40-402, Protection of Human Subjectsin Biomedical and Behavioral Research.

This convenience sample initially included 32 (11 female and 21 male)healthy subjects with myopia (−1.00 to −8.75 spherical equivalent)between the ages of 21 and 46 years. At this center, all patients with refractiveconditions who were previously treated or who were being treated had similarage distribution and preoperative refractive error of −1.00 to −8.50spherical equivalent. However, at the end, these data were analyzed and measuredfor 30 patients. One subject's visual acuity was not collected on postoperativeday 3 because the patient was unable to keep the appointment. The second subjectwas dropped from the analysis because of his or her failure to simply recordhis or her daily subjective pain scores for all 3 postoperative days. Allsubjects were active-duty military members. Each subject had a bilateral consecutivelaser treatment consisting of LASEK performed in one eye and PRK performedin the contralateral eye following the first laser treatment at the same sitting.All surgical procedures were done by the same surgeon (A.P.). The order ofthe surgical procedures (ie, right eye first or left eye first) and the choiceof procedures (ie, PRK in the right eye and LASEK in the left eye or LASEKin the right eye and PRK in the left eye) was determined in advance usinga block randomization table. The first eye was randomized to either PRK orLASEK and the contralateral eye was automatically given the other laser procedure.This method of assignment to groups was used to help control for the possibleconfounding effects of surgery order and eye order on the main outcome measures.In addition, subjects were not told which procedure would be performed oneach eye, thus making this study a double-masked clinical trial. Ametropiawas targeted for both eyes in all patients.

Tetracaine hydrochloride eyedrops were used for topical anesthesia preoperatively.Twenty percent alcohol was applied to the center of the cornea for 45 secondsusing a corneal well. It was then thoroughly rinsed with a balanced salt solution.An 8.5-mm epithelial trephine (Shahinian LASEK trephine; Katena Products Inc,Denville, NJ) was used to create a 270° epithelial hinge. The epithelialflap was reflected back from the cornea with an epithelial peeler (SloaneLASEK Micro Hoe; Katena Products Inc). Excimer laser treatment (VISX Star3; VISX Inc, Santa Clara, Calif) was performed using the same normogram asthat used for the PRK procedure by the same surgeon (A.P.). The epithelialflap was rinsed with a balanced salt solution before being repositioned withthe edges overlapping the original intact epithelium. One minute was allowedfor the readherence of the epithelium to the underlying stromal tissue beforea bandage contact lens was placed in the eye. Lubricating eyedrops were immediatelyinitiated and instillation continued every hour at minimum for the first 3days until epithelial healing was complete. An ophthalmic solution of 0.3%ofloxocin (Ocuflox; Allergan Inc, Irvine, Calif) eyedrops were started immediatelyafter the completion of the surgical procedure and instillation continued4 times daily for 3 days. A sterile ophthalmic solution of 0.5% ketorolactromethamine (Acular; Allergan Inc), 4 times daily, was given for the first24-hour postoperative period. After bandage contact lens removal, fluorometholonewas administered 4 times daily and continued for 1 month if the preoperativerefractive error was greater than −3.00. If the refractive error wasless than −3.00, fluorometholone therapy was tapered by 1 eyedrop perweek.

Tetracaine eyedrops were used for topical anesthesia as in the LASEKprocedure. The epithelium was removed using an automatic epithelial scrubber(Amoils; Advance Vision Technology Co Ltd, Bangkok, Thailand). The laser treatmentwas performed using an excimer laser (VISX Star 3; VISX Inc) by the same surgeon(A.P.). The eye was washed, thereafter, using a balanced salt solution. Abandage contact lens was placed in the eye. Treatment with 0.3% ofloxocin(Ocuflox; Allergan Inc) eyedrops was started immediately after the completionof the surgical procedure and continued 4 times daily for 3 days. A sterileophthalmic solution of 0.5% ketorolac tromethamine (Acular; Allergan Inc),4 times daily, was given for the first 24-hour postoperative period. Afterbandage contact lens removal, fluorometholone was administered 4 times dailyand continued for 1 month if the preoperative refractive error was greaterthan −3.00. If the refractive error was less than −3.00, fluorometholonetherapy was tapered by 1 drop per week.

An optometrist (S.N.) evaluated all measurements; each (the optometristand the patient) was unaware of the surgical procedure performed in each eye.For the subjective pain scores, patients were given the Faces Pain Scale andasked to rate the pain level in each of their eyes on a scale from 1 (leastpainful) to 10 (most painful) on postoperative days 1 through 3. Patientswere asked to return the scale by the end of their first postoperative weekvisit. Visual acuity was evaluated on postoperative days 1, 3, 7, and 30 usingSnellen visual acuity charts. To evaluate change in spherical equivalent,measurements were taken both preoperatively and 1 month postoperatively. Finally,to evaluate corneal epithelial healing, epithelial defects were measured onpostoperative days 1, 3, and 7; total surface area of the defect was reportedin square millimeters. Patients were also asked about the subjective qualityof their vision postoperatively and their preference of laser treatment.

Statistical significance for all analyses was set at P<.05. Most data were not normally distributed; for this reason,medians and percentiles were used to describe these data; nonparametric techniques(Mann-Whitney test, 2-tailed) were used to compare the surgical groups (PRKvs LASEK) at each of the different time points. The change in spherical equivalentdata was normally distributed; therefore, a t testwas used to determine any difference between the surgical groups, and 95%confidence intervals are reported.

The results of the Mann-Whitney test indicated that there was no differencein pain scale scores between patients with PRK-treated eyes and patients withLASEK-treated eyes on days 1 (z score = −0.29, P = .77), 2 (z score = −1.37, P = .17), or 3 (z score = −0.05, P = .96). The results are summarized in Table 1.

The results of the Mann-Whitney test indicated that there was no differencein uncorrected visual acuity between patients with PRK-treated eyes and patientswith LASEK-treated eyes on days 3 (z score = −0.48, P = .63), 7 (z score = −0.17, P = .86), or 30 (z score = −0.49, P = .62). The uncorrected visual acuity is reported asthe denominator of the Snellen visual acuity chart in Table 2.

To calculate change in spherical equivalent, preoperative measurementsof refractive error were subtracted from postoperative measurements; a positivechange value thus indicates an improvement in spherical equivalent and viceversa (Table 3). Results of a t test indicated no difference in the change in sphericalequivalent between the PRK-treated eyes and the LASEK-treated eyes (t = −0.38, P = .71). Inaddition, all spherical equivalent change scores were positive in both surgicalgroups, meaning that no patients exhibited worsened spherical equivalent postoperatively.

The area of epithelial defect (in square millimeters) was computed bymultiplying the horizontal (in millimeters) and vertical (in millimeters)diameter of the epithelial defects (Table4). On postoperative day 1 the results of the Mann-Whitney testindicated that patients with LASEK-treated eyes had a significantly smallerarea of epithelial defect compared with patients with PRK-treated eyes (z score = −4.91, P<.001).On postoperative day 3 the results of the Mann-Whitney test indicated thatpatients with PRK-treated eyes had a significantly smaller area of epithelialdefect compared with patients with LASEK-treated eyes (z score = −4.10, P<.001). On postoperativeday 7 statistical techniques were not used for these data because all patientshad an epithelial defect measurement of 0 regardless of surgical group.

Immediately after the completion of the consecutive bilateral procedure,patients were asked to point to the eye with better visual acuity. All patientspointed to the PRK-treated eye. These data were recorded on the patients'medical records.

At the end of 30 days, patients were asked the same question once again.Fifteen patients selected the PRK-treated eye and the other 15 selected theLASEK-treated eye. These data were recorded on the patients' medical records.

Immediately after the completion of the procedure patients were askedto point to the eye which they favored the laser treatment on, 15 patientspointed to their PRK-treatment eyes and the other 15 patients pointed to theirLASEK-treatment eyes. These data were recorded on the patients' medical records.

As reported in the medical literature, Camellin and Cimberle¹⁷ were among the first authors to describe and performthe LASEK procedure. Azar et al⁵ describeda similar but differently named surgical procedure about the same time. Thepopularity of LASEK has recently been gaining momentum among refractive ophthalmologistsafter a few clinical series showed that LASEK might have significant clinicaladvantages over PRK.³ Reported advantages includedthe reduced level of pain, faster visual recovery, and possible decreasedrisk for corneal haze formation.^2,5,7

Although previous studies have shown faster visual recovery and decreasedpain levels in patients with LASEK-treated eyes^5,7 noneof the studies have attempted to compare these factors within the same subjectreceiving different procedures in each eye in a double-masked format. To ourknowledge, our randomized prospective clinical trial is the first study ofthis kind which attempted to quantify the reported advantages of the LASEKprocedure by comparing it with the PRK procedure in the same subjects in adouble-masked technique.

In the review of the published literature, we noted that Camellin andCimberle¹⁷ reported a series of 249 LASEK-treatedcases with wide variability in the ease of making epithelial flaps. More thanhalf of their patients in the study felt no pain after the first day. In theirprospective comparative trial of 50 eyes with a follow-up of 1 month in whichthe results of LASEK and PRK were analyzed, Litwak et al¹⁴ reportedless discomfort and better visual acuity in the PRK-treated eyes during theearly postoperative period. Lee et al¹⁸ reportedin their prospective study of 84 bilateral consecutive LASEK-treated eyesfor a moderate myopia with a 6-month follow-up that LASEK is an effectiveand safe refractive procedure with a mean epithelial healing time of 3.68days. In another study, Lee et al¹⁵ prospectivelycompared the results of LASEK in 1 eye vs PRK in the other eye of 27 patients(54 eyes) with myopia (−3.00 to −6.50 spherical equivalent) performed2 weeks apart and reported lower postoperative pain and corneal haze scoresin the LASEK-treated eyes at 1 month with more than half (63%) of the participantspreferring the LASEK procedure. This study, although designed to be randomized,was not masked and the laser treatments were not done consecutively in a simultaneousfashion. Shahinian¹⁹ reported in his prospective12-month study of 146 consecutive LASEK-treated eyes that LASEK was safe andeffective in treating myopia and astigmatism with varying degrees of painand blurry vision for several days. Claringbold²⁰ reportedin his retrospective noncomparative interventional case series of 222 consecutiveLASEK-treated myopic eyes no loss of best-corrected Snellen visual acuityand no serious complications. Shah et al²¹ intheir prospective, nonrandomized, comparative-paired eye trial of 72 moderatemyopic eyes (36 patients) with a mean 1-year follow-up reported no statisticallysignificant difference between the 2 groups in their postoperative mean changein spherical equivalent. The corneal haze was also less in the LASEK-treatedgroup. However, this trial neither evaluated the patient's pain level northe epithelial healing rate.

In our study, we did not observe the differences reported in the literaturebetween LASEK- and PRK-treated eyes. There was no statistically significantdifference in postoperative visual acuity between the PRK- and the LASEK-treatedeyes, nor was there a difference in the subjective pain level between these2 groups. In addition, none of the eyes in either the PRK- or the LASEK-treatedgroup lost any lines of visual acuity compared with their best-corrected preoperativevisual acuity (ie, all spherical equivalent change scores were positive).However, the epithelial healing pattern was somewhat different between PRKand LASEK. On postoperative day 1, the median area of the epithelial defectin LASEK-treated eyes (1.0 mm²) was significantly smaller comparedwith the PRK-treated eyes (16.0 mm²). However, by postoperativeday 3, the median area of epithelial defect in the LASEK-treated eyes (4.0mm²) was significantly larger relative to the PRK-treated eyes(0.0 mm²). Indeed, while the median size of the epithelial defectdecreased from postoperative days 1 to 3 in the PRK-treated eyes (16.0-0.0mm²), it increased in the LASEK-treated eyes between days 1 and3 (1.0-4.0 mm²). We would like to call this phenomenon of nonadherenceand natural breakdown of the epithelium the early onsetpostoperative LASEK–epithelial degradation. It is possible thatepithelial flaps exposed to alcohol during the LASEK procedure lose theircellular integrity, thus, are unable to adhere and recover in the same fashionas the newly deposited and formed epithelium could do in the PRK-treated eyes.Nevertheless, the total days needed for the complete recovery of the corneawas similar by postoperative day 7 for both procedures. No epithelial defectswere detectable in either group of eyes. Furthermore, no corneal haze wasobserved for either of the treatment groups as was expected owing naturallyto the short follow-up time of this study. The corneal haze in PRK and LASEKprocedures typically occur 3 to 4 months afterward. No intraoperative or postoperativeflap complications, any other complications, or bad outcomes occurred in anyof our patients in the interval of this study.

About half of our patients preferred PRK over LASEK because of its quickerprocedural completion time; the PRK treatments took an average of 4 minutesto complete vs an average of 11 minutes for the LASEK procedure. On the otherhand, the other half of patients preferred LASEK because they did not likethe use of the epithelial scrubber on their eyes during the PRK procedure.However, overall the patients' satisfaction with their final postoperativevisual acuity was equal between the PRK and LASEK treatment after 30 days(data collected via self-report).

Another interesting observation was that, in general, the PRK-treatedeyes demonstrated better visual acuity and patients naturally preferred thePRK-treated eyes immediately after the completion of the procedure, comparedwith the LASEK-treated eyes, most likely because of the presence of epithelialedema (data collected via self-report). Patients nowadays have been told andhave come to expect that their visual acuity would be outstanding immediatelyafter undergoing a refractive procedure.³ Asa result, this immediate postoperative visual blurring could be disconcertingand disappointing for patients who have come to expect to have a perfect immediatepostoperative visual outcome.

In our study, we also noted intersubject variability and unpredictabilityin the ease of making epithelial flaps in the LASEK-treated eyes. Typicallywe noted that it was significantly more difficult to complete the epithelialflap in younger patients compared with older patients (aged >30 years).

After completing this study and comparing LASEK with PRK in an objectivemanner, we do not share the same enthusiasm expressed by our colleagues forthe superiority of LASEK over PRK. The LASEK procedure typically takes anaverage of twice as long for a surgeon to perform compared with the PRK procedure.Patients' anxiety may increase with increased surgical time. In addition,in the market-driven refractive practices of today, the longer time necessaryto perform the LASEK procedure, without its significant benefits, would adverselyaffect busy refractive surgical centers.

From our standpoint, additional studies are warranted to create epithelialflaps with different concentrations of alcohol or with different automatedmachines. Research on the long-term stability of visual acuity in a largeLASEK series could also provide additional information about the stabilityand efficacy of this procedure. One-year follow-up results of the visual acuity,refractive error, and corneal findings for both our PRK- and the LASEK-treatedpatients will be presented in the near future.

The previously described advantages of LASEK compared with PRK, suchas faster epithelial healing, shorter visual rehabilitation, and decreasedpostoperative pain compared with PRK were not observed in this study. Basedon our data, we see no additional benefits from the LASEK procedure comparedwith the PRK procedure in low to moderately myopic eyes.

Corresponding author and reprints: Amir Pirouzian, MD, DepartmentOf Ophthalmology, San Diego Children's Hospital, 3030 Children's Way, Suite109, San Diego, CA 92123-4282.

Submitted for publication December 4, 2002; final revision receivedJuly 23, 2003; accepted August 25, 2003.

This study was performed under US Air Force Surgeon General–approvedClinical Investigation No. FDG20020025H.

The views expressed herein are those of the authors and do not necessarilyreflect the official policy or position of the US Government, the US Departmentof Defense, or the US Department of the Air Force.