Sub-Bowman keratomileusis (SBK) is a type of LASIK

Transcription

1 Cornea Confocal Comparison of Corneal Nerve Regeneration and Keratocyte Reaction between FS-LASIK, OUP-SBK, and Conventional LASIK Fengju Zhang,*,1 Shijing Deng, 2 Ning Guo, 1 Mengmeng Wang, 1 and Xuguang Sun*,2 PURPOSE. We compared the regeneration characteristics of injured corneal subbasal nerve fibers (CSNFs) and keratocyte reaction following LASIK with a femtosecond laser (FS-LASIK), One Use-Plus Sub-Bowman s keratomileusis (OUP-SBK), and M2 90 microkeratome to provide a more suitable treatment strategy. METHODS. A total of 22 eyes that underwent FS-LASIK, 32 eyes that underwent OUP-SBK, and 26 eyes that underwent conventional LASIK were analyzed by confocal microscopy. Morphologic modifications of corneal architecture were evaluated postsurgically at 1, 3, and 6 months, and comparisons were made by ANOVA analysis. RESULTS. The density of stromal keratocytes in the FS-LASIK group was higher than that of OUP-SBK and conventional LASIK groups at 1 month postsurgery (P < 0.05), while in the OUP-SBK group it was higher than that of the FS-LASIK and conventional LASIK groups at 3 months postsurgery (P < 0.05). Regeneration of CSNFs was detected in 90% of eyes in all three groups around 3 mm of the peripheral flap at 1 month postoperation. CSNFs in 6.3% of the eyes in the OUP- SBK group were even observed around the peripheral flap at 6 mm 3 months after surgery, CSNFs were found in 62.5% of eyes (conventional LASIK), 72% of eyes (OUP-SBK), and 64.3% of eyes (FS-LASIK) around 3 mm near the center of the flap. While at 6 months, CSNFs were found in 100% of the eyes in all three groups. CONCLUSIONS. Keratocyte reactions after FS-LASIK and OUP-SBK were a little more severe than that after conventional LASIK. The repairing velocity of subbasal nerve fibers in the OUP-SBK group was a little faster than that of the FS and conventional LASIK groups. (Invest Ophthalmol Vis Sci. 2012;53: ) DOI: /iovs Sub-Bowman keratomileusis (SBK) is a type of LASIK procedure in which the produced corneal flap is significantly thinner centrally and more planar in shape, compared to outcomes achieved with the conventional LASIK approach. One of the main advantages of creating a thin corneal flap during SBK is the resulting sufficiently robust stromal tissue, which allows for safe excimer laser ablation, particularly in patients with moderate to high myopia who are at elevated risk for post-lasik keratectasia. 1,2 The femtosecond laser, which enables the creation of a lamellar corneal flap, provides refractive surgeons with a valuable alternative to mechanical flap creation. However, its commercialization in emerging nations is being inhibited by its steep costs. In light of that fact, alternatives procedures, such as SBK, using the One-Use Plus microkeratome system (Moria SA, Antony, France), are commercially and clinically more prevalent in China due to their cost-effectiveness and reproducible clinical results. 3 However, a significant drawback of using a mechanical microkeratome or femtosecond laser is the ensuing exposure and insult to the corneal subbasal nerve fibers (CSNF), which occurs during the overlay of the microkeratome knife or femtosecond laser photodisruptive process across the cornea to produce the flap. Subsequently, photoablation of the target site, using an excimer laser, destroys the exposed nerves in the stromal bed. These changes, which initially may induce corneal hypoesthesia lasting for several months postsurgery, have been emphasized clinically. 4,5 Using confocal microscopy examination, we attempted to evaluate changes to the corneal tissue, along with the healing and regeneration characteristics of the CSNF following SBK, achieved with either a mechanical microkeratome or a femtosecond laser device. To our knowledge, this is the first study to investigate the reconstructive characteristics and evaluate the regenerative velocity of the CSNFs in the integral cornea, following LASIK with a femtosecond laser or an automated microkeratome combined with excimer laser ablation. From the 1 Beijing Tongren Eye Center, Beijing Tongren Hospital, Capital Medical University, Beijing Ophthalmology & Visual Sciences Key Lab, Beijing, China; and 2 Beijing Institute of Ophthalmology, Beijing Tongren Hospital, Capital Medical University; Beijing, China. Submitted for publication October 12, 2011; revised February 19 and May 17, 2012; accepted July 2, Disclosure: F. Zhang, None;S. Deng, None;N. Guo, None;M. Wang, None;X. Sun, None *Each of the following is a corresponding author: Fengju Zhang, Beijing Tongren Eye Center, Beijing Tongren Hospital, Capital Medical University, Beijing Ophthalmology & Visual Sciences Key Lab, Beijing, China ; fjz269@yahoo.com. Xuguang Sun, Beijing Institute of Ophthalmology, Beijing Tongren Hospital, Capital Medical University, Beijing, China ; sunxuguang@yahoo.com. MATERIALS AND METHODS Patients We recruited 80 eyes of 40 myopic patients suitable for LASIK surgery in our refractive center through the routine pre-refractive surgery evaluation process. The average age was 23.7 years (range years), and the mean spherical equivalent (SE) was diopters (D; range 2.00 to 8.75 D). A detailed explanation of the procedure was offered to the patients before the selection of their preferred surgical choice. Patients were divided into three groups (depending on the flap). In group 1 (FS-LASIK group, 11 cases, 22 eyes, SE D) the flap was created by LASIK with a femtosecond laser (superior hinge, FEMTO LDV; Ziemer Ophthalmic Systems Group, Port, Switzerland). In group 2 (OUP-SBK group, 16 cases, 32 eyes, SE Investigative Ophthalmology & Visual Science, August 2012, Vol. 53, No Copyright 2012 The Association for Research in Vision and Ophthalmology, Inc.

2 IOVS, August 2012, Vol. 53, No. 9 Corneal Nerve Regeneration and Keratocyte Reaction 5537 FIGURE 1. Demarcation and score of the CSNF D) the flap was created by a nasal hinge One Use-Plus SBK automated microkeratome (Moria SA). In group 3 (13 cases, 26 eyes, SE D) the flap was created as in a conventional LASIK procedure (superior hinge, M2 Single-Use 90 automated microkeratome; Moria SA). The corneal stroma then was ablated without wavefront guidance using the VisX CustomVue Star S4 IR excimer laser machine (Abbott Medical Optics, Irvine, CA). Postsurgical eye care included antibiotics and corticosteroids (levofloxacin [Santen, Osaka, Japan] 4 times daily for 1 week, 0.1% fluorometholone [Allergan, Irvine, CA] 4 times daily for 1 week and gradually decreased for another 1 week) along with lubricant eye drops (4 times daily for 1 month), and was similar in all three groups. In addition to the routine follow-up examination, in vivo confocal microscopy (IVCM; HRT III; Heidelberg Engineering, Heidelberg, Germany) analysis was performed preoperatively and postoperatively at 1, 3, and 6 months. The mean preoperative pachymetry was 538 lm (range lm) measured by an SP-3000 ultrasound pachymeter (Tomey Corporation, Nagoya, Japan). The programmed corneal flap thickness was 110 lm for the FS-LASIK group, 110 lm for the OUP-SBK group, and 130 lm for the conventional LASIK group. For patients assigned to the automated microkeratome groups (i.e., OUP-SBK and conventional LASIK groups), one single-use OUP-SBK head and M2 Single-Use 90 head were used for each patient, and the selection of metallic suction rings was based on normograms provided by the manufacturer. The mean programmed diameter of the optical zone of all eyes was at least 6 mm without wavefront guidance. All patients were examined by the same operator (SD) with the HRT III/Rostock Cornea Module IVCM. In this prospective case-control study, patients in all three groups underwent uneventful refractive surgery with no intraoperative or postoperative complications, as assessed by repeated slit-lamp examinations and refraction measurements. With the exception of the TABLE. Evaluation of Cell Densities in the Surgical Interface for All Three Groups from 1 6 Months Postoperatively OUP-SBK Group (cells/mm 2 ) FS-LASIK Group (cells/mm 2 ) Conventional LASIK Group (cells/mm 2 ) 1 Month * Months * Months * P < 0.05.

3 5538 Zhang et al. IOVS, August 2012, Vol. 53, No. 9 FIGURE 2. The changes of keratocytes density at 1 month (A, D, G), 3 months (B, E, H), and 6 months (C, F, I) after surgery. (A C) Conventional group. (D F) OUP-SBK group. (G I) FS-LASIK group. Magnification, lm.

4 IOVS, August 2012, Vol. 53, No. 9 Corneal Nerve Regeneration and Keratocyte Reaction 5539 difference in average attempted corneal flap thickness and very high diopter refraction, all groups were fully comparable. All HRT III images were analyzed in a masked manner by one examiner (SD) to compare corneal patterns in each of three studied groups, with particular attention paid to the flap and stromal aspects, including keratocyte apparent density (calculating only the cell density of the corneal stroma in the surgical interface layer) and nerve regeneration in different corneal regions. Regenerative Velocity and Score Calculations The regenerative velocity and scores of the CSNFs were calculated in the following manner: The regeneration of CSNFs in the 4 quadrants and central part of the corneal flaps was observed using IVCM. During evaluation of the CSNFs, all 4 quadrants were examined and the patient was asked to focus on the contralateral fixation point corresponding to the quadrant the researcher wished to examine. Each subject was scanned from the margin of the corneal flaps to the anterior pole of the cornea. The nerve score of the longest regenerative CSNF was the respective score of the quadrant area, and the highest score of the 4 quadrants served as the score for the eye. Nerve regenerative velocity score for the cornea was divided into 4 concentric circle zones. The score was determined as 3 once the new CSNFs reached into the central corneal round zone of the 3 mm radius, 2 when the CSNFs reached into the circle zone for approximately 3 6 mm from the anterior pole, 1 when the CSNFs reached into the circle zone, in which radius is from 6 mm to approximately the margin of the flap, and 0 when no CSNF were found in the cornea flap (Fig. 1). We modified the method of Lagali et al. 6 and the keratocyte density was determined with the mean value of three images selected from the most anterior keratocyte layer of the ablated region (identified by the high reflected particles appearing between the keratocyte in the image) at three different part of central cornea, which could indicate the activation of keratocyte directly. Only cells with clearly distinguishable borders were counted. Cells crossing the top and right edges of the rectangular region were counted, while those crossing the bottom and left edges were excluded. The acquired two-dimensional image was defined by pixels covering an area of lm with a lateral digital resolution of 1 lm/pixel. Corneal morphology results were compared after LASIK surgery in all central and peripheral corneal layers, with particular attention paid to flap interface and flap margin areas (Fig. 1). Because the HRT III confocal microscope is a minimally invasive technique, informed consent was obtained from all participants after explanation of the nature of the study. All participants were treated in accordance with the tenants of the Declaration of Helsinki. Statistical Analysis The histologic characteristic of the new CSNFs was observed, and the nerve regenerative velocity scores of each group were compared. All score data were analyzed using the v 2 test, which was performed using SPSS for Windows 11.5 (SPSS Inc., Chicago, IL). ANOVA analysis was used to compare the data among the three groups. A P value of < 0.05 was accepted as statistically significant. FIGURE 3. The flap margin (arrow in arrow) of FS-LASIK (A) and OUP- SBK (B) groups appeared clear-cut edge. It was less clear in the routine LASIK group (C). Magnification, lm. RESULTS The results of flap thickness for all three groups were analyzed by IVCM. Preoperative central pachymetry was lm (FS-LASIK), lm (OUP-SBK), and lm (conventional LASIK). After surgery, the obtained central flap thickness examined by IVCM was lm (FS-LASIK), lm (OUP-SBK), and lm (conventional LASIK). There was a significant difference between the conventional LASIK group and the others (P < 0.001), but there was no significant

5 5540 Zhang et al. IOVS, August 2012, Vol. 53, No. 9 FIGURE 4. Changes of CSNFs in central cornea pre- (A) and postoperatively (B D) in the OUP-SBK group. The CSNFs were absent at 1 (B) and 3 (C) months in the central cornea after surgery. The regenerated CSNFs were appeared at 6 months (D) in the central cornea with a thinner, nonbranching and tortuous appearance (arrow). (E) The recovery of CSNF in the incision (arrow in the flap) at 1 month postoperatively. HRT III image magnification, lm. difference between the OUP-SBK and FS-LASIK groups (P > 0.05). Changing in corneal stroma cell density appeared in the flap interface layer from 1 to 6 months postsurgery among all three groups depicting keratocyte transformation, most likely related to cellular activation beneath the LASIK flap interface. Cell density results for all three groups are summarized in the Table and Figures 2A 2I. While at 1 month after surgery the density of the stromal keratocyte cells in the FS-LASIK group was higher than that of the OUP-SBK and conventional LASIK groups (P < 0.05), at 3 months after surgery it was highest in the OUP-SBK group compared to the FS-LASIK and conventional LASIK groups (P < 0.05). By 6 months postsurgery, there was no significant difference among the three groups (P > 0.05). The flap margin after the FS-LASIK and OUP-SBK techniques appeared microscopically as a clear-cut edge. This cell reaction diminished with time, leaving little fibrotic scar adjacent to a wound constriction originating from the surrounding stroma. The flap margin of the conventional LASIK procedure had the appearance of a less clearly identified fibrotic scar with no epithelial plug (Figs. 3A C) The regeneration of subbasal nerve fibers occurred one month after the surgery. Regeneration was detected in 90% of eyes in all three groups around 3 mm of the peripheral flap; in 6.3% of the eyes in OUP-SBK group it was even observed at 6 mm around the peripheral flap. At 3 months after surgery, CSNFs were found in 62.5% of eyes (conventional LASIK), 72% of eyes (OUP-SBK), and 64.3% of eyes (FS-LASIK) around 3 mm near the center of the flap, and in 100% of eyes in all three groups at 6 months (Figs. 4 6). Regenerated CSNFs in the three groups were derived mainly into the peripheral circle area from the edge of the flap incision and towards the corneal apex. Corneal innervation in the center of the flap disappeared, and subbasal nerve plexus was absent 1 month after surgery, but regeneration occurred from the peripheral cornea, passing through the cutting edge towards the center, as a thinner, nonbranching appearance (Fig. 4E). At 3 months postoperatively, intensively curved nerves covered one-third of the corneal flap, around the flap as tortuous and disordered

6 IOVS, August 2012, Vol. 53, No. 9 Corneal Nerve Regeneration and Keratocyte Reaction 5541 FIGURE 5. Changes of CSNFs in central cornea preoperatively (A) and postoperatively (B D) in the FS-LASIK group. The CSNFs were absent at 1 month (B) after surgery. The regenerated CSNFs appeared thinner (white arrow) at 3 months. More regenerated fibers appeared, but the density of regenerated CSNFs still decreased and the fibers were more tortuous (black arrow) than preoperatively (D). HRT III image magnification, lm. shapes located in the second circle area of approximately 6 3 mm distance to corneal apex, and in the corneal central 3 mm area from 3 6 months. The nerve regeneration was derived vertically from the CSNF stump in the incision and directly from the deep stroma. The CSNFs in the flap hinge still were fewer and narrower at 3 months postoperatively than preoperatively in the FS-LASIK group (Fig. 7). DISCUSSION Many studies in the past few years have reported a decrease in the number of subbasal nerve fibers and keratocytes after LASIK. 5,7 These changes initially may induce corneal hypoesthesia lasting for months. Recent studies involving the impairment and reconstruction of corneal nerves after LASIK surgery have focused only on the nerve fibers of the central part of the cornea. 7 9 In our study, the regenerative velocity of CSNFs in the integral cornea was observed in vivo from the margin of the corneal flaps to the anterior pole of the cornea by using IVCM up to 6 months postsurgery. The cornea was divided into 4 concentric circle zones to determine the nerve score. We then evaluated the nerve regenerative velocity by the scores. We propose that this method is objective, effortless, and time conserving, and merited to be used widely in the analysis and semi-quantification of the impairment and reconstruction of integral corneal nerves. It has been found that the femtosecond laser could trigger directly energy-induced keratocyte necrosis in the corneal stroma, and microkeratomes primarily trigger keratocyte apoptosis in the first few hours and up to a day following surgery. 10 Since the calculation of cell density is from the interface beneath the corneal flap after LASIK, and not the basal and stromal cells in the other different layers, the cell density from the surgical interface area could indicate the activity of inflammation. Therefore, there is no need to obtain the level cell density in the flap interface area preoperatively. There was no difference between the FS-LASIK and OUP-SBK groups in the flap thickness calculated by confocal microscopy; however, in both groups the flaps were significantly thinner than that of the conventional LASIK group. The results of flap depth were similar to the clinical data from OCT and

7 5542 Zhang et al. IOVS, August 2012, Vol. 53, No. 9 FIGURE 6. Changes of CSNFs in central cornea preoperatively (A) and postoperatively (B D) in the conventional LASIK group. The CSNFs were absent at 1 (B) and 3 (C) months in the central cornea after surgery. The regenerated CSNFs appeared at 6 months (D) in the central cornea with a thinner and more tortuous appearance (arrow). HRT III image magnification, lm. pachymetry despite the discrepancy during the in vivo confocal microscopy examination. The stromal tissues ablated by excimer laser in the OUP and FS groups were much more than that of the conventional group according to the refraction preoperatively. Our results indicated that the thinner the corneal flap was made by femtosecond laser or microkeratome, the stronger the keratocyte reacted to the injury. Therefore, the corneal stroma cell density in the flap interface layer in the FS-LASIK and OUP-SBK groups was higher compared to the conventional LASIK group 1 3 months after surgery in this study. However, why are there differences between the FS and OUP-LASIK groups at 1 and 3 months? This may be due to the small sample data and large standard deviation in these two groups, or possibly these differences are due to different tissue reactions caused by the femtosecond laser photo disruption or the microkeratome separation. This hypothesis may be confirmed further by larger sample data. The results suggested that the corneal cell reaction of wound healing was related to the lamellar cutting thickness and ablation depth during LASIK. Steroid eye drops in the FS- LASIK and OUP-SBK groups should be used for a longer period, compared to the conventional LASIK group. Kauffmann et al. 9 found that rarefied subepithelial nerve fibers were visible at the edges of the corneal flaps at 8 weeks after LASIK, and single nonbranched nerve fibers could be visualized in the center of the ablation zone after 3 months. There was no other report on the regeneration manner of the new CSNFs, except concerning the density and quantity of the nerve fibers in the central cornea. Lee et al. 7,10 and De Medeiros et al. 11 demonstrated that CSNFs lost during LASIK regenerated slowly and were visible until six months after surgery. Patel et al. observed that the CSNF density did not return to preoperative level even after 12 months. 12 The recovered CSNFs in the central cornea were visible only in 60% of postoperative eyes at 3 years after LASIK, observed by Calvillo et al. 5 These reports showed that the recovery of CSNFs in the central cornea is very slow, probably due to the injury caused by the traditional microkeratome, which results in a thicker flap compared to the modern modified microkeratome and thinner flap in our study. In our study, most of the nerve regeneration occurred from the CSNF stump in the

8 IOVS, August 2012, Vol. 53, No. 9 Corneal Nerve Regeneration and Keratocyte Reaction 5543 corneal incision; however, some of them could have originated directly from the deep stroma (only found in 5 patients whose data were not included in the score evaluation). The regeneration of subbasal nerve fibers occurred one month after the surgery, detected in 90% of eyes in all three groups around the peripheral flap 3 mm (score 1). These branchless fibers appeared tortuous and thin, and derived towards the corneal apex. At 3 months after surgery, score 3 was evaluated in 62.5% of eyes (conventional LASIK), 72.0% of eyes (OUP- SBK), and 64.3% of eyes (FS-LASIK), while score 3 was found in 100% of eyes (all three groups) at 6 months after surgery, which suggested CSNFs repairing in OUP-SBK was a little faster than that of the FS-LASIK and conventional groups. The reasons were analyzed as follows: Firstly, the deeper the corneal flap was cut, the more severe the damage was to nerves of the main branch, so the early regenerative velocity of CSNFs in conventional LASIK was slower than that of FS-LASIK and OUP-SBK groups. Secondly, in the FS-LASIK group, after exposure to the femtosecond laser, the corneal flap must be separated bluntly and then ablated with an excimer laser. Therefore, it has been questioned whether that additional mechanical flap dissection could induce extra injury for the CSNFs or not. This point should be investigated further besides the reason of femtosecond laser photodisruptive process. Thirdly, a nasal hinge may be better than a superior hinge comparing the regenerative velocity of CSNFs due to remaining more nerve branches according to the anatomy of the corneal nerve. The CSNFs density in the corneal flap hinge also was reduced in each group after surgery. At 1 month, these fibers appeared to be tortuous and tiny, and some even were interrupted. The density at 3 months was better than that at 1 month; however, it did not return to the preoperative level, which suggested that the hinge of the corneal flap also should be protected carefully during the surgery to restore the function of the ocular surface early. Our study showed that OUP-SBK and FS-LASIK were clinically safe and effective, especially for cases of high myopia and thinner cornea, and reduced occurrence of ectasia. As for the FS-LASIK, the procedure for corneal flap blunt separation should be delicate and careful, and the hinge also should be protected during laser ablation to avoid additional injury to the CSNF. While the OUP-SBK procedure may be more suitable or attractive to developing nations, allowing for satisfactory results and lower costs compared to FS-LASIK, further and longer term studies should be conducted to shed more light on this matter. References FIGURE 7. Changes of CSNFs in the hinge of the superior corneal flap preoperatively (A), and 1 (B) and 3 (C) months postoperatively. The density of CSNFs in the hinge decreased significantly and the branches became thinner at 1 month (white arrow). The CSNFs in the flap still are fewer and thinner at 3 months than preoperatively. HRT III image magnification, lm. 1. Condon PI, O Keefe M, Binder PS. Long-term results of laser in situ keratomileusis for high myopia: risk for ectasia. J Cataract Refract Surg. 2007;33: Randleman JB. Post-laser in-situ keratomileusis ectasia: current understanding and future directions. Curr Opin Ophthalmol. 2006;17: Lian JC, Zhang SS, Ye S, Dong SQ. Correlation analysis and corneal flap thickness changes among Moria SBK, 90 and 110 microkeratome in laser in situ keratomileusis. Chin Ophthalmic Res. 2010;28: Stachs O, Zhivov R, Kraak R, Hovakimyan M, Wree A, Guthoff R. Structural-functional correlations of corneal innervation after LASIK and penetrating keratoplasty. J Refract Surg. 2010; 26: Calvillo MP, McLaren JW, Hodge DO, Bourne WM. Corneal reinnervation after LASIK: prospective 3-year longitudinal study. Invest Ophthalmol Vis Sci. 2004;45:

9 5544 Zhang et al. IOVS, August 2012, Vol. 53, No Lagali N, Germundsson J, Fagerholm P. The role of Bowman s layer in corneal regeneration after phototherapeutic keratectomy: a prospective study using in vivo confocal microscopy. Invest Ophthalmol Vis Sci. 2009;50: Lee SJ, Kim JK, Seo KY, Kim EK, Lee HK. Comparison of corneal nerve regeneration and sensitivity between LASIK and laser epithelial keratomileusis (LASEK). Am J Ophthalmol. 2006;141: Moilanen JA, Holopainen JM, Vesaluoma MH, Tervo TM. Corneal recovery after lasik for high myopia: a 2-year prospective confocal microscopic study. Br J Ophthalmol. 2008;92: Kauffmann T, Bodanowitz S, Hesse L, Kroll P. Corneal reinnervation after photorefractive keratectomy and laser in situ keratomileusis: an in vivo study with a confocal videomicroscope. Ger J Ophthalmol. 1996;5: Lee BH, McLaren JW, Erie JC, Hodge DO, Bourne WM. Reinnervation in the cornea after LASIK. Invest Ophthalmol Vis Sci. 2002;43: De Medeiros FW, Kaur H, Agrawal V, et al. Effect of femtosecond laser energy level on corneal stromal cell death and inflammation. J Refract Surg. 2009;25: Patel SV, McLaren JW, Kittleson KM, Bourne WM. Subbasal nerve density and corneal sensitivity after laser in situ keratomileusis: femtosecond laser vs mechanical microkeratome. Arch Ophthalmol. 2010;128: