November 2016




Review of “1-year experience in myopia correction with transepithelial PRK case- matched with femtosecond-assisted LASIK”

by James O’Brien, MD, Ilya Sluch, MD, Andrew Melson, MD, Joshua Hendrix, MD, Jason Smart, MD, Christina Lippe, MD, Vincent Venincasa, MD, Kyle Rogers, MD, Russell Burks, MD, Clifton Fay, MD, Andrew Rodenburg, MD, and Valerie Lobodiak, MD, Dean McGee Eye Institute residents


David Jackson, MD
David W. Jackson, MD, Clinical associate professor, Dean McGee Eye Institute

How does transepithelial PRK compare to femtosecond laser-assisted LASIK? I asked the Dean McGee (Oklahoma) residents to review this head-to-head study that appears in this month’s issue of JCRS.

–David F. Chang, MD, EyeWorld journal club editor

From left: Kyle Rogers, MD, Russell Burks, MD, James O’Brien, MD, Clifton Fay, MD, Andrew Rodenburg, MD, Jason Smart, MD, Christina Lippe, MD, Joshua Hendrix, MD, Valerie Lobodiak, MD, Andrew Melson, MD, Ilya Sluch, MD, Vincent Venincasa, MD, and R. Michael Siatkowski, MD Source: Dean McGee Eye Institute

Since the first application of photorefractive keratectomy (PRK) on human eyes in 1987 and laser in-situ keratomileusis (LASIK) in 1991, the field of refractive surgery has blended and modified the two procedures in order to combine the safety profile of PRK with the recovery time of LASIK. With the advent of more precise laser technologies, we are now able to offer transepithelial photorefractive keratectomy (transPRK), femtosecond-assisted LASIK (femtoLASIK), femtosecond lenticule extraction (FLEx), and small incision lenticule extraction (SMILE). TransPRK treatment parameters account for the variance in epithelial ablation, and it delivers a single photoablation treatment to the cornea without the need for epithelial removal or flap creation. Meanwhile, the femtoLASIK platform utilizes a 1053 nm femtosecond laser to create a bladeless corneal flap, followed by stromal ablation, which avoids the complications of a microkeratome. While all of these procedures have data supporting their safety and efficacy, few studies compare them head to head. This particular study compares 1-year postoperative refractive outcomes of transPRK versus femtoLASIK.

Study summary

This study was a retrospective, case-matched comparative series that evaluated the long-term postoperative visual outcomes of transPRK and femtoLASIK refractive surgery techniques. Ninety-eight patients (196 eyes) who underwent transPRK were compared to 196 case-matched eyes that underwent femtoLASIK. TransPRK patients were matched with femtoLASIK patients by calculating the differential blur vector, a representation of the patients’ spherical, cylindrical, and axis components of their refractive error. All patients had baseline examination including both uncorrected and corrected distance visual acuity, manifest refraction, corneal topography and pachymetry, pupillometry, slit lamp examination, and dilated fundus examination. On postoperative day 7, uncorrected and corrected distance visual acuity and manifest refraction were recorded, followed by 3- and 12-month postoperative examinations that included all of the components of the baseline assessment. The outcomes of primary interest were the postoperative visual acuity and refractive error.

The procedures were performed by 11 different surgeons in Utrecht, Netherlands. Patient preference, surgeon preference, and corneal thickness were factors that determined whether transPRK or femto LASIK was chosen for a patient. Patients were excluded if they were “systemically ill,” or had postoperative residual stromal bed thickness less than 300 microns, preoperative corneal thickness less than 470 microns, or “abnormal” corneal topography. Retreatment cases were also excluded. Systemic illness and abnormal corneal topography were not further defined.

The average age of patients treated with transPRK was 36±11 years (range 18 to 79 years) and 39±10 years (range 20 to 62 years) for patients who underwent femtoLASIK (p<0.01). The two treatment groups had similar preoperative mean spherical refractive error (–3.36±1.78 D and –3.34±1.74 D for transPRK and femtoLASIK, respectively, p=0.2). There was a significant difference with astigmatic error (mean 1.10±0.91 D and 1.00±0.84 D for transPRK and femtoLASIK, respectively, p<0.0001), with corresponding difference in mean topography between groups (43.82±1.47 D and 43.47±1.38 D for transPRK and femtoLASIK, respectively, p<0.01). Both groups were similar with respect to uncorrected (p=0.1) and corrected (p=0.4) preoperative visual acuities, refraction stability, and contact lens holidays. At the day 7 postoperative examination, there was a significant difference in uncorrected (0.14±0.16 and 0.03±0.13 logMAR for transPRK and femtoLASIK, respectively, p=0.0001) and corrected (0.09±0.12 and –0.06±0.06 logMAR for transPRK and femtoLASIK, respectively, p=0.0001) distance visual acuities between groups. There was also a significant difference in mean spherical equivalent (0.04±0.47 D and –0.12±0.52 D for transPRK and femtoLASIK, respectively, p=0.002) and mean astigmatic (0.19±0.41 D and 0.29±0.34 D for transPRK and femtoLASIK, respectively, p=0.006) refractive error between groups. At the 3-month postoperative examination, there was no significant difference between the groups in mean uncorrected (–0.01±0.15 and –0.01±0.12 logMAR for transPRK and femtoLASIK, respectively, p=0.4) or mean corrected (–0.08±0.08 and –0.08±0.06 logMAR for transPRK and femtoLASIK, respectively, p=0.4) visual acuities. However, a significant difference in mean spherical equivalent (+0.11±0.51 D and –0.08 ±0.40 D for transPRK and femtoLASIK, respectively, p=0.0001) and mean astigmatic (0.34±0.45 D and 0.27±0.30 D for transPRK and femtoLASIK, respectively, p=0.03) refractive error was present. Lastly, at the 12-month postoperative examination, there was a significant difference in mean uncorrected (–0.05±0.14 and –0.01±0.14 logMAR for transPRK and femtoLASIK, respectively, p=0.005) and mean corrected (–0.11±0.06 and –0.09±0.05 logMAR for transPRK and femtoLASIK, respectively, p=0.03) visual acuities as well as mean spherical equivalent (0.11±0.56 D and –0.09±0.46 D for transPRK and femtoLASIK, respectively, p=0.0001) refractive error between the two groups. There was no significant difference in astigmatic error. The achieved versus attempted refractive correction outcomes were also assessed. There was no significant difference in spherical equivalent correction between the groups. There was, however, 8% overcorrection of astigmatism in the transPRK group and 1% undercorrection of astigmatism in the femtoLASIK group. Overall, 83% of the transPRK eyes and 85% of the femtoLASIK eyes were within 0.5 D of their target refractive goals. Both treatment groups demonstrated overall stability in their postoperative visual and refractive outcome during the follow-up period. The uncorrected visual acuity stabilized earlier for the femtoLASIK group (3 months) than the transPRK group (up to 12 months). Similarly, the postoperative astigmatic error in the femto LASIK group stabilized between the 7-day and 12-month follow-up, whereas the transPRK group stabilized by 3 months.

In the discussion, the authors comment on the higher variance noted in transPRK patients, which they attributed to the use of population-based epithelial profiles rather than individualized, patient-specific profiles. Despite this variance, however, there was not a clinically significant difference in final visual acuity outcome for transPRK patients at 12 months compared to the femtoLASIK group. They concluded that transPRK seemed to feature comparable if not better long-term visual outcomes than femtoLASIK, but that femtoLASIK featured faster recovery times. Despite these differences, however, they concluded that neither procedure can be considered superior based on this single retrospective study.


While a prospective, case-matched controlled study or a randomized clinical trial would be more desired, the authors attempted to limit bias of their retrospective comparative series design by using a consecutive series of transPRK patients and matching them with the femtoLASIK population during the same time period. While the average age of transPRK group was statistically significantly lower than that of the femtoLASIK group, this is unlikely to have any clinical significance. Decision for transPRK was based on patient and surgeon preference as well as corneal thickness, which adds an element of selection bias, especially when 11 surgeons were included in the study. However, the use of two identical laser units does help in normalizing the sample. The exclusion criteria used by the authors helps to generalize the study results to healthy refractive surgery candidates.

Interestingly, corneal topographic and thickness analysis was performed solely with the Nidek OPD-Scan Placido disc topographer (Nidek, Gamagori, Japan) and ultrasound pachymetry, and did not include Scheimpflug imaging or Belin-Ambrosio ectasia testing. All treatments for transPRK and femtoLASIK were performed with the same SCHWIND ORK-CAM planning module (SCHWIND eye-tech-solutions, Kleinostheim, Germany), further reducing bias in outcomes. Treatment protocols were similar between the two groups, correcting for manifest refraction. The main differences were the extra wetting of the corneal epithelium with a balanced salt solution-soaked sponge for even wetting of the corneal surface and use of 0.02% mitomycin-C for 30 seconds in the transPRK group. Postoperatively, the transPRK group received topical corticosteroids only after epithelial closure and was treated with fluorometholone taper over a 9-week period, while the femtoLASIK group was tapered off of steroids over a 2-week period. The relatively hyperopic outcomes in the transPRK group (0.11±0.56 D) versus femtoLASIK (–0.09±0.46 D) could possibly be explained by a longer course of steroids inducing less myopic regression. Better vision at the day 7 postoperative visit would be expected in the femtoLASIK group due to faster healing time. The study included only patients achieving 1-year follow-up, which is appropriate, and presented excellent and stable refractive outcomes. There was a slightly better visual outcome in the transPRK group, possibly due to the slight overcorrection, which was statistically significant.

One important limitation of the statistical methods used in the study is that the inter-group comparisons did not account for within-patient correlation. This is important as more than half of cases were bilateral. Since all of the transPRK eyes were bilateral, taking one eye from a bilateral case to analyze, although alleviating the issue of within-patient correlation, would decrease the sample size by 50% for this group. The authors performed a large number of comparisons using t-tests and Chi-square tests, thus inflating the chance of false positive findings. Procedures for multiple comparisons such as Bonferroni or Dunnett should be considered at least for comparisons of primary interest. Given that the data is longitudinal, analyses based on mixed models will generally be more powerful than using many t-tests. Other important demographic and/or clinical factors (e.g., corneal thickness, keratometry, astigmatism, etc.) may need to be adjusted (if significant) as well. Additionally, because the procedures were performed by 11 different surgeons, it may be helpful to perform the subgroup analysis, stratified by surgeon, in order to examine the potential bias in the final results.

In summary, this study suggests similar refractive outcomes between transPRK and femtoLASIK. However, the study did not quantify postoperative pain, dry eye symptoms, patient satisfaction, or willingness to repeat the procedure, all of which are important when comparing procedures of different pain levels, corneal penetration, and recovery times. It would also be interesting to perform age-stratified subgroup analysis to see if there is a difference in outcomes between younger and older patients as there was a wide age range among the included patients. While the sample size was sufficient to measure refractive outcomes, a much larger sample size would be necessary to evaluate the complication rate since the incidence of corneal haze and flap dislocation has decreased in recent years with improved surgical technique. The cost differential between the two procedures will also likely be an important factor to consider in the future. A large prospective study comparing transPRK, femtoLASIK, and SMILE with broader outcome metrics is desired to elucidate whether one procedure is superior to the others.


1. Chen LY, et al. Comparison of femtosecond and excimer laser platforms available for corneal refractive surgery. Curr Opin Ophthalmol. 2016;27:316–322.

2. Haq Z, et al. Infections after refractive surgery. Curr Opin Ophthalmol. 2016;27: 367–372.

3. dos Santos AM, et al. Femtosecond laser-assisted LASIK flap complications. J Refract Surg. 2016;32:52–59.

4. Li SM, et al. Laser-assisted subepithelial keratectomy (LASEK) versus photorefractive keratectomy (PRK) for correction of myopia. Cochrane Database Syst Rev. 2016;2:CD009799.

Editors’ note: The faculty mentors/reviewers were David Jackson, MD, clinical associate professor, and R. Michael Siatkowski, MD, residency program director, Dean McGee Eye Institute.

Contact information


Related articles:

Femto in cataract surgery: The state of the technology by Liz Hillman EyeWorld Staff Writer

Large study to compare conventional phaco to femtosecond laser surgery by Erin L. Boyle EyeWorld Senior Staff Writer

Dissonance in a femtosecond by J.C. Noreika, MD, MBA

A new twist on femtosecond platforms by Maxine Lipner EyeWorld Senior Contributing Writer

An update on billing policy for femtosecond laser use by Ellen Stodola EyeWorld Staff Writer

1-year experience in myopia correction 1-year experience in myopia correction
Ophthalmology News - EyeWorld Magazine
283 110
220 125
True, 11