Femtosecond (FS) laser vision correction procedure for moderate to high myopia: a prospective study of ReLEx® flex and comparison with a retrospective study of FS-laser in situ keratomileusis

Authors


Jesper Ø. Hjortdal, MD, PhD
Department of Ophthalmology
Aarhus University Hospital
Nørrebrogade 44
8000 Aarhus C
Denmark
Tel: + 45 89493222
Fax: + 45 86121653
Email: vestergaard_anders@hotmail.com

Abstract.

Purpose:  To present our initial clinical experience with ReLEx® flex (ReLEx) for moderate to high myopia. We compare efficacy, safety and corneal higher-order aberrations after ReLEx with femtosecond laser in situ keratomileusis (FS-LASIK).

Methods:  Prospective study of ReLEx compared with a retrospective study of FS-LASIK. ReLEx is a new keratorefractive procedure, where a stromal lenticule is cut by a femtosecond laser and manually extracted. Forty patients were treated with ReLEx on both eyes. A comparable group of 41 FS-LASIK patients were retrospectively identified. Visual acuity, spherical equivalent (SE) and corneal tomography were measured before and 3 months after surgery.

Results:  Preoperative SE averaged −7.50 ± 1.16 D (ReLEx) and −7.32 ± 1.09 D (FS-LASIK). For all eyes, mean corrected distance visual acuity remained unchanged in both groups. For eyes with emmetropia as target refraction, 41% of ReLEx and 61% of FS-LASIK eyes had an uncorrected distance visual acuity of logMAR ≤ 0.10 at day 1 after surgery, increasing to, respectively, 88% and 69% at 3 months. Mean SE was −0.06 ± 0.35 D 3 months after ReLEx and −0.53 ± 0.60 D after FS-LASIK. The proportion of eyes within ±1.00 D after 3 months was 100% (ReLEx) and 85% (FS-LASIK). For a 6.0-mm pupil, corneal spherical aberrations increased significantly less in ReLEx than FS-LASIK eyes.

Conclusions:  ReLEx is an all-in-one femtosecond laser refractive procedure, and in this study, results were comparable to FS-LASIK. Refractive predictability and corneal aberrations at 3 months seemed better than or equal to FS-LASIK, whereas visual recovery after ReLEx was slower.

Introduction

For a decade, the femtosecond laser has been used in refractive surgery. The near-infrared laser pulses create photodisruption, to cut corneal tissue with minimal collateral tissue damage (Slade 2007). Until recently, femtosecond lasers have primarily been used as an alternative to the microkeratome to cut thinner and planar corneal flaps in laser in situ keratomileusis (LASIK). Laser in situ keratomileusis as treatment of myopia is considered one of the most successful surgical procedures overall, with a patient satisfaction rate of 95% (Ang et al. 2009). Despite this overall success of LASIK procedures, the two-step approach with flap cutting and subsequent excimer laser ablation can potentially be improved, by techniques minimizing the changes in stromal hydration. Hence, all corneal refractive surgical procedures affect corneal hydration, and this effect needs to be limited, as the amount of ablated tissue is dependent on stromal hydration (Kim & Jo 2001; Patel et al. 2008). Also, well-known risk factors as epithelial ingrowth and loose flaps need to be limited. Recently, femtosecond lasers have become available for intrastromal lenticule cutting and subsequent lenticule extraction, the so-called refractive lenticule extraction or ReLEx® (Sekundo et al. 2008). ReLEx seems to be comparable to femtosecond-based LASIK (FS-LASIK), although clinical comparisons have not yet been published (Blum & Sekundo 2010). Theoretically, this new laser technique has the potential to minimize the aforementioned challenges of LASIK.

Two types of ReLEx procedures have been described, ReLEx® flex and ReLEx smile. In ReLEx® flex (femtosecond lenticule extraction), the intrastromal lenticule is cut below a hinged flap as in a FS-LASIK procedure (Blum et al. 2010; Sekundo et al. 2008); whereas in ReLEx smile (small-incision lenticule extraction), the incision site has been minimized, and the procedure is without a flap (Sekundo et al. 2011; Shah et al. 2011). Both types of ReLEx procedures seem safe and promising in terms of the refractive correction of myopia, although ReLEx smile is not yet widely performed (Blum et al. 2010; Sekundo et al. 2008, 2011; Shah et al. 2011).

The purpose of this article is to present our initial clinical experience with ReLEx® flex (ReLEx) for moderate to high myopia, in a prospective study. We then compare the results with a retrospective study of FS-LASIK.

Materials and Methods

Subjects and pre-examination

During the period from March to June 2010, the first 50 patients (100 eyes) underwent ReLEx for myopia at our department. Patients had myopia ranging from −6.00 to −10.25 dioptres (spherical equivalent) and astigmatism ranging from 0 to 2.25 dioptres (see Table 1 for baseline biometric parameters). Mean age was 37.8 ± 8.4 years and range 22–54 years, and 33% were men. To better evaluate the outcome of the new ReLEx procedure, a comparable group of our first 50 patients treated with FS-LASIK during the period from June to September 2009 was identified retrospectively. Here, mean age was 36.8 ± 7.4 years, range 24–52 years, and 22% were men.

Table 1.   Results at baseline and 3-month follow-up.
 ReLExFS-LASIKp-valueReLExFS-LASIKp-value
BaselineBaseline3 months3 months
  1. CDVA = corrected distance visual acuity, CCT = central corneal thickness, FS-LASIK = femtosecond LASIK, UDVA = uncorrected distance visual acuity.

  2. p-values are determined by two-tailed t-test if distributed normally or else non-parametric tests, between ReLEx and FS-LASIK. For proportions, Fisher’s exact test was used. CCT was measured at pupil centre with Pentacam HR.

  3. * Here, only eyes with emmetropia as their target refraction were included, that is, n = 34 of 40 ReLEx eyes and n = 36 of 41 FS-LASIK eyes.

Average keratometry (D)
 Mean ± SD44.03 ± 1.3543.91 ± 1.360.60538.36 ± 1.5638.71 ± 1.830.417
 Range41.69–46.8140.71–46.7535.58–41.9533.92–43.49
CCT (μm)
 Mean ± SD547 ± 30558 ± 280.073439 ± 29471 ± 31<0.001
 Range486–640510–638393–521415–575
Eyes with UDVA ≤ 0.10 logMAR88%*69%*0.081*
CDVA (logMAR)
 Mean ± SD0.00 ± 0.050.00 ± 0.050.6760.00 ± 0.080.01 ± 0.060.653
 Range−0.10 to 0.10−0.10 to 0.10−0.10 to 0.30−0.10 to 0.10
Change in CDVA (number of eyes)
 Lost 2 lines200.241
 Lost 1 line580.547
 Unchanged21240.658
 Gained 1 line1290.455
Spherical equivalent (D)
 Mean ± SD−7.50 ± 1.16−7.32 ± 1.090.809−0.06 ± 0.35*−0.53 ± 0.60*<0.001*
 Range−6.00 to −10.25−5.00 to −9.75−0.75 to 0.50*−1.75 to 0.50*
Difference in attempted versus achieved spherical equivalent (D)
 Mean ± SD0.00 ± 0.41−0.45 ± 0.62<0.001
 Range−0.75 to 0.88−1.75 to 0.88
Within ±0.50 D of target refraction83%56%0.016
Within ±1.00 D of target refraction100%85%0.026

Patients underwent a thorough eye examination including objective and manifest visual acuity, intraocular pressure (NIDEK TONOREF II, Gamagori, Japan), pupil size (NIDEK pupillometer, Gamagori, Japan), keratometric measurements, slit-lamp examination and fundoscopy. They had to discontinue contact lenses for 2 days (soft lenses) and 2 weeks (hard lenses) prior to assessment.

Inclusion criteria were as follows: a minimum age of 20 years, stable myopia for at least 1 year, corrected distance visual acuity (CDVA) ≥20/25 (≥0.80) and no other ocular diseases or conditions except myopia with or without astigmatism. Pregnancy or breastfeeding meant exclusion from surgery. Regular topographic patterns of both the corneal front and back were confirmed with a Pentacam HR Scheimpflug camera (Oculus, Germany). This included the usage of the Pentacam Ambrósio/Belin module to exclude subclinical keratoconus. Furthermore, the central corneal thickness (CCT) was measured with both an OLCR pachymeter (Haag-Streit, Koeniz Switzerland) (Ivarsen et al. 2009) and a Pentacam HR tomograph. Only eyes with a calculated postoperative residual stromal bed of more than 250 μm were included. To avoid discrepancy between expectations and actual outcome after surgery, all patients were explained the risks and potential outcomes of refractive surgery, and subsequently, informed consent was obtained from all participants.

ReLEx and FS-LASIK procedures

All surgical procedures were performed by two experienced surgeons (JH and SA). The desired refractive change was entered directly into the laser. A VisuMax® femtosecond laser (Carl Zeiss Meditec, Jena, Germany) was used for ReLEx treatments and FS-LASIK flaps. In FS-LASIK, subsequent photoablation was performed with a MEL-80 excimer laser (Carl Zeiss Meditec). Both procedures are Conformité Européenne (CE) marked. The ReLEx and the FS-LASIK procedures were performed bilaterally and under topical anaesthesia using two drops of oxybuprocaine tetrachloride applied 5 and 1 min before surgery. The patient was positioned under the curved contact glass of the VisuMax femtosecond laser and asked to fixate a blinking target. When appropriate centration (centre of pupil) was observed, suction was applied to the contact glass. Femtosecond laser pulses (500 kHz, 150 nJ) were focused in a spiral pattern with a spot distance of 3.0 μm, causing the tissue to be cleaved. The ReLEx procedure is illustrated in Fig. 1. First, the laser created the back of the intrastromal lenticule, with photodisruption from the periphery to the centre of the cornea. Then, the laser created the lenticule front with an anterior lamellar cut (centre → periphery) that was extended towards the surface to create a flap with a superiorly placed hinge. Flap diameter was 8 mm with a cut angle of 90° and a lenticule diameter of 6.5 mm. Intended flap thickness was 120 μm in all cases. After laser treatment, a thin, blunt spatula was used to break the remaining tissue bridges, and the flap was lifted. The stromal lenticule was loosened with a spatula, grasped with a pair of forceps and removed. The flap was repositioned, and the corneal surface and interface was flushed with saline. This ReLEx procedure is also known as ReLEx® flex. The surgical technique has previously been described in detail (Blum et al. 2010; Sekundo et al. 2008). In FS-LASIK, only the flap was created with the 500 kHz femtosecond laser. Here the settings were laser energy with a range of 100–200 nJ, spot distance 3.5 μm, a flap diameter of 8 mm, a cut angle of 70° and an intended flap thickness of 120 μm. The excimer laser ablation was performed using smart ablation algorithm in a 6.0-mm-diameter zone. At the end of the procedure, the cornea was irrigated with saline in both procedures, and all patients received one drop of chloramphenicol and one drop of Voltaren® Ophtha (Novartis Healthcare, Copenhagen, Denmark) (Diclofenac). The postoperative regime included Flurolon® (Fluormetholon, Allergan Pharmaceuticals, County Mayo, Ireland) and chloramphenicol drops four times a day for 1 week, followed by two times a day for 1 week. The patients were encouraged to use lubricating drops as needed to improve comfort and facilitate surface repair.

Figure 1.

 Schematic illustration of the ReLEx® flex procedure. 1: Corneal top and side view; the VisuMax femtosecond laser creates the back and the front of the intrastromal lenticule. The lamellar cut is then extended peripherally towards the surface to create a flap with a superiorly placed hinge. 2: After laser treatment, the flap is lifted and the stromal lenticule is removed. 3: The flap is repositioned.

Outcome measures

Follow-up appointments after ReLEx were scheduled at day 1, 1 week, 1 and 3 months, and after FS-LASIK at day 1 and after 3 months. The following parameters were obtained at each visit: corrected (CDVA) and uncorrected distance visual acuity (UDVA) using Snellen charts, objective and manifest refraction, CCT, keratometry, Pentacam HR tomography, tonometry (not on day 1) and a slit-lamp examination to evaluate the anterior segment. Complications, including interface scatter, debris and flap folds, were noted. Patients, who missed their 3-month follow-up appointment, were contacted by phone or letter and given a new appointment. After two missed appointments, the patient was considered lost to follow-up.

Visual acuity from the Snellen chart was converted to logMAR units using the Holladay method (Holladay 1997).

Total (anterior + posterior) corneal wavefront error (root mean square, μm) was calculated from Pentacam tomography data over the 4- and 6-mm-diameter central corneal zone. Wavefront aberrations were decomposed into Zernike polynomials up to the 8th order. Standard Zernike notation with double indexing was used, as previously defined (Hjortdal et al. 2002; Thibos et al. 2002). In this study, we focused on three of the most clinically important higher-order aberrations (HOAs): spherical aberration, coma and trefoil. Wavefront aberrations were based on the following refractive indices: cornea = 1.376 and aqueous = 1.336.

Statistics

All measured data were collected and entered into the patients’ journal and in an anonymous form into a standardized study spreadsheet in Microsoft Excel 2007. Statistical analyses were performed using Microsoft Excel and systat sigmaplot 12 (Systat Software Inc., San Jose, CA, USA). Statistical analysis for visual acuity was based on logMAR units. As most of the outcome data were not normally distributed, the Mann–Whitney (Wilcoxon) rank sum test was used to compare the mean outcomes for non-normally distributed data. If data were normally distributed, two-tailed Student’s t-test was used. For repeated measurements, the paired Student’s t-test and Wilcoxon signed rank test were used. The Kolmogorov–Smirnov test in SigmaPlot was used to test for normality. p < 0.05 was considered statistical significant. As the two eyes of one subject are not independent, only one eye of each subject was used in the comparative analyses. Eyes were randomly chosen using random number generator in Excel 2007.

Ethics

The study was conducted in agreement with the tenets of the Declaration of Helsinki. According to Danish healthcare law, this study was considered a quality control study. Therefore, ethical approval was not needed. However, all participants were thoroughly informed at the preoperative examination and gave oral consent before surgery.

Results

Fifty patients (100 eyes) were treated with the ReLEx procedure. Forty of these (80%) attended the 3-month follow-up. Forty-one of 50 patients were included in the FS-LASIK group. The remaining nine patients were subsequently excluded because of failure to fulfil all inclusion criteria, or no 3-month follow-up data existed (patients were considered lost to follow-up).

There were no statistically significant differences (p > 0.050) in baseline age and gender in the two groups. Also, preoperative mean spherical equivalent (SE), astigmatism, mean CDVA, mean keratometric readings, mean CCT and corneal aberrations were similar in the two groups (p > 0.050). See Tables 1 and 2 for results.

Table 2.   Tomography-derived corneal wavefront errors after ReLEx/FS-LASIK.
 ReLExFS-LASIKp-valueReLExFS-LASIKp-value
BaselineBaseline3 months3 months
  1. HOA = higher-order aberrations, SA = spherical aberrations, FS-LASIK = femtosecond laser in situ keratomileusis.

  2. Values are given as mean ± standard deviation. p-values are determined by two-tailed t-test if distributed normally or else Mann–Whitney rank sum test, between ReLEx and FS-LASIK.

  3. * Marks p-values below 0.05.

  4. inline image, inline image, inline image.

HOA – 4 mm pupil
 SA (RMS, μm)0.047 ± 0.0310.036 ± 0.0200.0690.094 ± 0.0760.091 ± 0.0490.838
 Coma (RMS, μm)0.070 ± 0.0520.067 ± 0.0430.8610.202 ± 0.2740.142 ± 0.0790.247
 Trefoil (RMS, μm)0.074 ± 0.0450.081 ± 0.0460.5370.156 ± 0.1040.124 ± 0.0610.092
HOA – 6 mm pupil
 SA (RMS, μm)0.227 ± 0.0790.217 ± 0.0860.6100.442 ± 0.2340.700 ± 0.153<0.001*
 Coma (RMS, μm)0.193 ± 0.1420.197 ± 0.1250.6470.459 ± 0.3680.538 ± 0.3020.075
 Trefoil (RMS, μm)0.050 ± 0.0270.064 ± 0.0460.3380.118 ± 0.0760.098 ± 0.0480.482

Uncorrected distance visual acuity

After surgery, UDVA improved significantly in both groups. Thirty-four ReLEx eyes and 36 FS-LASIK eyes had emmetropia as target refraction. At day 1, a significant difference in favour of FS-LASIK was found with 41% of ReLEx eyes and 61% of FS-LASIK eyes having an UDVA of ≥20/25 (logMAR ≤ 0.1; p = 0.033). However, after 3 months, it was with 88% for ReLEx and 69% for FS-LASIK (p = 0.030; Table 1; Fig. 2).

Figure 2.

 Efficacy. Cumulative percentage of eyes attaining specified cumulative levels of uncorrected distance visual acuity 3 months after surgery. Only eyes with emmetropia as target refraction were included.

Corrected distance visual acuity and safety

Corrected distance visual acuity did not change significantly in either of the two groups (p > 0.050; Table 1). The proportion of eyes with CDVA ≥20/20 (logMAR ≤ 0.00) for ReLEx was 85% both before and after surgery. Corresponding figures for FS-LASIK were 88% and 83%. One ReLEx patient had a central corneal abrasion during surgery and developed mild intrastromal haze. Another eye developed trace haze because of difficulty loosening the intrastromal lenticule during operation. Both eyes lost two lines of CDVA. A total of two ReLEx eyes and no FS-LASIK eye had lost two lines of CDVA at the 3-month follow-up (Fig. 3). When including both eyes of all patients in the analysis, 3 of 80 ReLEx eyes and 1 of 82 FS-LASIK eye had lost two lines of CDVA at the 3-month follow-up. All eyes had a CDVA ≥20/40 (logMAR ≤ 0.30) 3 months after surgery. No signs of postoperative ectasia were noted, neither clinically nor tomographically. No other complications during or after surgery were noted.

Figure 3.

 Safety. The percentage of eyes in which there was a gain/loss of lines in corrected distance visual acuity 3 months after surgery. All eyes in both groups were included.

Predictability, efficacy and stability

Both ReLEx and FS-LASIK (Fig. 4) proved efficient at achieving the attempted correction 3 months after surgery. However, a significantly greater proportion of ReLEx patients were within ±0.5 and ±1.0 D of the target SE refraction, as compared to FS-LASIK patients (Table 1). As not all eyes had emmetropia as target refraction, the mean SE for all patients in the two groups could not be directly compared. However, comparing the difference in attempted versus achieved correction for the two groups, a significant difference (p < 0.001) in favour of ReLEx was found at the 3-month follow-up (Table 1; Fig. 5). The same result was found when comparing the 34 ReLEx eyes and 36 FS-LASIK eyes that had emmetropia as their target refraction The ReLEx procedure showed stability in postoperative refraction from day 1 to 3 months after surgery (Table 3).

Figure 4.

 Predictability. Scatter plot and linear regression analysis of the attempted SE refractive change plotted against the achieved SE refractive change at 3 months (SE = spherical equivalent).

Figure 5.

 Predictability. Spherical equivalent refractive accuracy at 3-month follow-up. The percentage of eyes attaining specified differences in attempted versus achieved correction for the two groups. All eyes in both groups were included. (SE = spherical equivalent).

Table 3.   Stability for all ReLEx eyes.
 Mean SE (D) ± SDNo. of eyes
  1. Mean spherical equivalent (SE) ± standard deviation from preoperative to 3 months after surgery.

Preoperative−7.50 ± 1.1640
Day 1−0.09 ± 0.5840
1 week−0.09 ± 0.4740
1 month−0.13 ± 0.5338
3 months−0.18 ± 0.5540

Central corneal thickness measurements (Table 1) showed a significantly greater amount of tissue removal in ReLEx eyes (p < 0.001) because of a larger treatment zone (6.5 mm in ReLEx compared with 6.0 mm in FS-LASIK). The average decrease in CCT was 108 ± 15 μm for ReLEx and 87 ± 16 μm for FS-LASIK eyes. For all ReLEx eyes, this resulted in an average of 14-μm corneal tissue removed pr. spherical equivalent. For all FS-LASIK eyes, it was 12 μm/SE.

Higher-order aberrations

In both groups, 3rd, 4th and other HOAs increased after treatment (Table 2). Three months after surgery, no significant differences in corneal HOAs were observed between the groups at the 4-mm-diameter zone. However, there were significant differences between the groups at the 6-mm-diameter zone where ReLEx showed less spherical aberration (p < 0.001).

Discussion

Femtosecond lasers have led to renewed interest in lamellar refractive surgery, with their ability to cut corneal tissue with high precision, and with the possibility of making minimal invasive surgery. Also, the problem with variations in stromal hydration has renewed the interest in refractive surgery without the use of an excimer laser. Theoretically, a short-time, low-suction, one-step femtosecond laser procedure, with optimal energy settings, and with the refractive cut made before exposing the stromal tissue, could minimize hydration variations and maybe improve refractive predictability. Also, minimal invasive surgery could potentially reduce the risk of ectasia and traumatically loosened flaps.

When comparing UDVA at day 1 after surgery, we found that FS-LASIK eyes recovered faster than ReLEx eyes. As previously described (Blum et al. 2010), slit-lamp examination showed a transiently enhanced visibility of the interface in some ReLEx eyes during the first postoperative week. It is speculated that this type of interface scatter after cutting of the intrastromal lenticule is to blame for the slightly reduced visual acuity the first day after surgery, and we speculate that suboptimal cutting of the stromal fibres is the reason (Heichel et al. 2011). Further optimizing laser energy, spot size and spacing could potentially create a more ‘smooth stromal cut’ and minimize this delayed recovery in the future. After 1 week, uncorrected visual acuity in ReLEx eyes had improved to at least the same level as FS-LASIK eyes at day 1. Uncorrected distance visual acuity in FS-LASIK eyes remained almost constant from day 1 to 3 months postoperatively, while in contrast, ReLEx eyes improved significantly in UDVA from both day 1 to 1 week and from 1 week to 3 months. Thus, after 3 months, UDVA was significantly better in the group of ReLEx eyes. When comparing proportion of eyes with an UDVA of ≥20/25 (logMAR ≤ 0.1) at the 3-month follow-up, we found slightly better results than in Blum’s study of ReLEx® flex (Blum et al. 2010). However, in Blum’s study, most of the patients were undercorrected and not treated towards emmetropia. Direct comparison is therefore not possible. In a study concerning FS-LASIK, Blum et al. (2009) found that 97% (31/32) of the operated eyes had an UDVA of ≥20/25 (logMAR ≤ 0.1) at the 3-month follow-up. This surpasses our results, but again, the study had fewer eyes and with a lower degree of myopia than the eyes included in our study. Shah et al. (2011) found that 79% of all full-corrected eyes had an UDVA of ≥20/25 (logMAR ≤ 0.1) with the ReLEx smile method after 6 months. Overall, in this study, the ReLEx procedure seemed equal to or better than the FS-LASIK procedure with regard to postoperative refractive predictability. However, visual acuity showed faster recovery after FS-LASIK than after ReLEx.

Concerning CDVA and safety, no episodes of suction loss or an excessive amount of opaque bubble layer were noted in this study. In one eye, there was substantial difficulty of loosening the lenticule, and it meant that more force was used to remove the lenticule causing an abrasion. This eye later developed microstria and trace haze. The induced corneal abrasion healed within 1–2 days after surgery. Two ReLEx eyes and no FS-LASIK eyes lost two lines of CDVA at 3-month follow-up (Fig. 3). Despite this, both ReLEx patients were satisfied with the surgery and wished no further follow-up. There was no significant difference between the two groups (Table 1). Blum reported 2 (of 107) eyes who lost two or more lines after 3 months, with one resolving to no loss of CDVA after prolonged steroid treatment (Blum et al. 2010). Sekundo et al. (2011) reported 2 (of 91) eyes with a loss of two or more lines of CDVA after 6 months. Shah et al. (2011) observed no eyes with a loss of two lines of CDVA, but one episode of steroid-induced glaucoma. We did not experience any episodes of intraocular pressure rise or noted signs of diffuse lamellar keratitis. The 10 patients lost to follow-up were all contacted by either telephone or mail and offered a new 3-month appointment. When contacted, no one had any complaints. Overall, the ReLEx procedure seemed as safe as FS-LASIK.

Concerning predictability, a significant difference in favour of ReLEx was found concerning the 3-month postoperative mean SE refraction, when comparing eyes with emmetropia as target (Table 1). ReLEx eyes were closer to emmetropia than the FS-LASIK eyes. Chan et al. (2008) found a mean SE refraction of −0.20 D at the 3-month follow-up, but in a group of patients with a lower degree of myopia, than ours (FS-LASIK). At 6-month follow-up, Blum et al. (2010) found a mean SE of −0.19 ± 0.47 D (ReLEx® flex), Sekundo et al. (2011) found −0.01 ± 0.49 D (ReLEx smile) and Shah et al. (2011) found +0.03 ± 0.30 D (ReLEx smile). Again, these trials included eyes with a smaller mean preoperative SE refraction than compared with our clinical study. We also found a tendency towards slight undercorrection in the FS-LASIK group (Figs 4 and 5). If re-treatment is needed, the ReLEx technique could prove difficult to perform, for low degrees of myopia or hyperopia. Thus, perhaps an excimer-based enhancement procedure after lifting the flap or a photorefractive keratectomy (PRK) procedure should be performed. No ReLEx patients in this study needed re-treatment.

Higher-order aberrations commonly increase after LASIK procedures. Several studies report less induction of HOAs with the use of femtosecond lasers rather than microkeratomes during LASIK (Buzzonetti et al. 2008; Medeiros et al. 2007; Tran et al. 2005). We found a significant increase in spherical aberrations (SA), coma and trefoil in both groups and a significant difference in the induction of SA for 6-mm pupils after 3 months. This difference in HOAs after ReLEx and FS-LASIK may in part be due to the difference in treatment zone diameter between the two treatment modalities. Thus, slight decentration of the 6-mm-diameter ablation zone in FS-LASIK may cause part of the edge to be included in the 6-mm-diameter corneal analysis, increasing the amount of HOAs. In contrast, the wavefront analysis would be expected to be within the 6.5-mm treatment zone in ReLEx. Still, the fundamentally different approach to refractive surgery – lenticule extraction compared with excimer photoablation – may also be a possible explanation for the difference in HOAs between the two groups. The previously mentioned problems with corneal hydration could also be speculated as another reason for differences in HOAs. In Sekundo’s initial study of ReLEx® flex (Sekundo et al. 2008), they did not find significant differences pre- and postoperatively in ocular wavefront values. However, their sample size was small (10 eyes), and they also used a 5.0-mm analysis diameter compared with our 4.0- and 6.0-diameter analyses. For the ReLEx smile technique, Shah et al. (2011) found a significant increase in postoperative total HOAs over a 5.0-mm-diameter analysis. Both studies measured whole-eye aberrations as compared to our corneal tomography-derived wavefront measurements. Overall, in this study, the ReLEx procedure seemed comparable to or better than FS-LASIK in minimizing the induction of HOAs.

There are some limitations to our study: First, this was a retrospective analysis with regard to the FS-LASIK group. In this process, there was a risk of bias, and significant differences between the prospective and the retrospective results must therefore be viewed with caution. Also, this could maybe explain our partly inferior results compared with other published results concerning FS-LASIK. Optimally, both techniques should be compared in prospective studies with a longer follow-up. Second, the cut angles at the flap edge, the spot sizes and the laser energy settings were slightly different in the two groups. However, this is not expected to play a major role for our results. Third, 10 ReLEx patients failed to complete the 3-month follow-up, and also, nine FS-LASIK patients were not included in the retrospective analysis. These patients did not seem to differ from the rest of the two groups.

In conclusion, we report our first clinical experience with the ReLEx® flex technique in a prospective study. For comparison, a matched group of patients with FS-LASIK were retrospectively reviewed. After 3 months, refractive predictability was better for ReLEx compared with FS-LASIK-treated eyes. CDVA was unchanged in both groups. Femtosecond-LASIK eyes recovered faster than ReLEx eyes, in terms of uncorrected visual acuity. This was probably due to interface scatter during the first week in ReLEx eyes. Total corneal HOAs increased in both groups, but were equal to or less in eyes treated with lenticule extraction compared with photoablation. Safety seemed high and comparable. Further studies, especially prospective studies with a longer follow-up period, are needed to evaluate this new technique. A further development of the ReLEx technique with a smaller incision also seems promising in terms of corneal refractive correction of myopia.

Acknowledgements

Pentacam HR was donated by Bagenkop Nielsens Myopi-Foundation. Commercial Relationship: Hjortdal J.: Travel reimbursement. Other authors: None. Results concerning visual acuity, predictability and corneal aberrations were presented at ARVO May 2011, Ft. Lauderdale, Florida, and at ESCRS September 2011, Vienna, Austria.

Ancillary