Reinnervation in the Cornea after LASIK

Abstract

purpose. Nerve fibers in the cornea are disrupted by photorefractive procedures. In this study, the denervation and reinnervation of human central corneas were evaluated by sequential, quantitative measurements of nerves viewed by confocal microscopy in vivo during the first year after LASIK.

methods. Seventeen eyes were studied of 11 patients who had undergone LASIK to correct myopia from −2.0 D to −11.0 D. Eyes were treated with an excimer laser with a planned 180-μm flap. Central corneas were scanned throughout their full thicknesses by confocal microscopy before and at 1 week and 1, 3, 6, and 12 months after LASIK. Nerve fiber bundles appeared as bright, well-defined, linear structures that were sometimes branched and usually appeared in several consecutive frames. The number of nerve fiber bundles per scan in two to eight scans per eye per visit was determined in the subbasal region, the full-thickness stroma, the stromal flap (layer between the most anterior keratocyte and the flap interface), and the stromal bed (layer between the flap interface and the endothelium).

results. In the subbasal region, the number of nerve fiber bundles decreased by more than 90% 1 week after LASIK and was significantly lower at all times after surgery than it was before surgery (P < 0.001). It increased 6 and 12 months after LASIK, but remained less than half of the preoperative value. In the stromal flap, the number of nerves at all times after surgery was also significantly less than before surgery (P < 0.001) and did not increase significantly by 1 year. In the stromal bed, there were no significant differences among any of the nerve measurements before and after LASIK (P = 0.24).

conclusions. In the corneal flap, the number of subbasal and stromal nerve fiber bundles decreases by 90% immediately after LASIK. During the first year after LASIK, subbasal nerve fiber bundles gradually return, although by 1 year their number remains less than half of that before LASIK.

The cornea is richly innervated by nerve fibers of the ophthalmic division of the trigeminal nerve. ¹ Owing to this dense sensory innervation, the cornea is extremely sensitive to external stimuli, a property that is very important for stimulating tear secretion, maintaining normal physiologic balance, and initiating aversion reflexes that serve to protect the eye. ²  

Keratorefractive surgery disrupts the integrity of corneal nerves. In photorefractive keratectomy (PRK) the epithelium and anterior stroma, which contain most of the nerve fibers, are removed. In laser in situ keratomileusis (LASIK), which is now the more widely used procedure for correcting myopia, the microkeratome cuts the subbasal nerve bundles and the superficial stromal nerve bundles in the flap interface, although nerves that enter the flap through the hinge region are spared. The nerves of the anterior stromal bed, which is subsequently ablated, are destroyed by the excimer laser treatment. Nerve fibers gradually reinnervate the cornea and sensation slowly returns, ^{3 4 5 6} although it is not known how quickly the nerves return or whether their number is as high as it was before treatment. 

Subbasal and stromal nerves have been studied histologically in postmortem specimens obtained after LASIK in rabbits ⁷ and in specimens obtained after LASIK in humans. ^{8 9} Clinical confocal microscopy has provided a means of repeated noninvasive examination of corneal nerves. ^{10 11} The clinical confocal microscope has been used to examine corneal nerves after LASIK procedures in humans in both cross-sectional ^{3 12} and longitudinal ¹³ studies. None of the investigations, however, attempted to quantify the number of nerves or their density. In this longitudinal study, we used confocal microscopy to measure sequential changes in subbasal and stromal nerve density for 12 months after LASIK. 

Methods

Patients

The study included 17 eyes of 11 patients (the remaining 5 eyes in these patients had reoperations for undercorrections and were excluded). There were 10 women and 1 man, aged 20 to 46 years (mean age ± SD, 32.2 ± 9.2). All eyes had normal anterior segments, intraocular pressures (≤22 mm Hg), and fundi. Contact lens wear was discontinued 2 weeks (soft lenses) or 3 weeks (hard lenses) before the operation. Patients who had diabetes mellitus or glaucoma or were currently using any topical ocular medication were excluded from the study. The mean preoperative spheroequivalent of refraction was −6.56 ± 2.44 D (range, −2.00 to −11.00 D). No patients wore contact lenses after LASIK. The Mayo Clinic Institutional Review Board approved the study, which conformed to the principles of the Declaration of Helsinki for research involving human subjects. All patients provided informed consent after the nature and possible consequences of the study were explained to them. 

Surgical Procedure and Postoperative Treatment

Myopia or myopic astigmatism was corrected by LASIK performed with an excimer laser (Star; VISX, Santa Ana, CA). A flap was created with a planned thickness of 180 μm by using a microkeratome (Hansatome; Chiron Vision Corp., Claremont, CA). Patients fixated on a target during the ablation. The mean planned ablation depth was 62.8 ± 26.2 μm (range, 18–110). The stromal bed was irrigated with balanced salt solution at room temperature before and after flap replacement to eliminate residual debris. At the end of the operation, the flap was allowed to dry in place for at least 3 minutes to facilitate adhesion. 

After LASIK, topical medications consisted of fluorometholone 0.1% (FLM; Allergan Inc., Irvine, CA) four times per day for 1 week and tapered over 2 weeks, and ofloxacin 0.3% (Ocuflox; Allergan Inc., Irvine, CA) four times per day for 5 days. 

Confocal Microscopy

Corneas were examined by using a tandem scanning confocal microscope (Tandem Scanning Corp., Reston, VA) before surgery and at 1, 3, 6, and 12 months after LASIK. The method of examination was described in an earlier publication, ¹⁴ and a brief description is given herein. The microscope objective lens was cleaned with 70% isopropyl alcohol wipes before and after each examination. A drop of 2.5% hydroxypropyl methylcellulose (Goniosol; CIBA Vision Ophthalmics, Atlanta, GA) was placed on the tip of the objective lens as an optical coupling medium, and the lens was manually advanced until the medium contacted the surface of the central cornea. A full-thickness scan, consisting of a series of confocal images, was recorded as the focal plane was advanced at approximately 78 μm/sec from anterior to the epithelium to posterior to the endothelium. Digital images were stored on a computer workstation (Indy; Silicon Graphics, Inc., Mountain View, CA) at 30 frames/sec. Each image represented a coronal section approximately 475 × 350 μm (horizontal × vertical) and was separated from adjacent images by approximately 2.6 μm. A full-thickness scan required 7 to 9 seconds to complete. On each visit, the cornea was scanned through its full thickness four to eight times. The objective lens and coupling medium were separated from the cornea between each scan and then reapplied. Scans were not in the identical region each time, but all scans were within the central 4 mm of the cornea. Thus, the scans represented random, full-thickness samples of the central cornea. 

Nerve Fiber Bundles

All confocal scans of sufficient quality for nerve visualization were evaluated, and the scan position (corneal depth) of each nerve fiber bundle was recorded. Nerves appeared as long, narrow structures (Fig. 1) and those longer than 50 μm were counted. If the length of the visible portion of a nerve, including its appearance in adjacent scans, was less than 200 μm, the length was recorded. Nerve branches were not counted (a branched nerve was considered a single nerve). The number of nerve fiber bundles located in the subbasal region (several frames anterior to the most anterior keratocyte), in the full-thickness stroma, in the stromal flap (distance from the most anterior keratocyte to the flap interface), and in the stromal bed (distance from the flap interface to the endothelium) was recorded as the mean number per scan for each cornea for each examination (Fig. 2) . The appearance of small, bright objects (presumably metal particles) in the anterior stroma was used to identify the interface in postoperative corneas. The thicknesses of the stromal flap and the stromal bed determined from the 1-month post-LASIK scans were used to delimit the corresponding anterior and posterior stromal layers in the preoperative cornea. One observer (BHL) evaluated all scans, except for the 1-week postoperative scans, which were evaluated by a second observer (JCE). Adequate scans were available (two to eight scans per eye per visit) for all 17 eyes for all preoperative and postoperative examinations. 

Statistical Analysis

The postoperative nerve fiber bundle data were not normally distributed, primarily because of the large number of scans without visible nerves. Therefore, they were summarized by using medians and interquartile ranges (25th and 75th percentiles, Q25 and Q75, respectively). Median numbers of nerve fiber bundles in each region at each observation time were compared by using the Friedman test, a nonparametric version of the repeated measures analysis of variance. Significant differences were investigated after adjusting for multiple comparisons by using the Student-Newman-Keuls procedure. P < 0.05 was considered statistically significant. Generalized estimating equation (GEE) models were completed to account for any potential correlation between the two eyes of an individual. In all cases, results of the GEE model were similar to results of the Friedman test, and only the results of the Friedman test are presented. 

Results

The numbers of nerve fiber bundles in each region of the cornea at each observation time are given in Table 1 and Figure 3 . Of the approximately 6000 nerve fiber bundles identified in this study, only 43 were less than 200 μm in visible length. They were all in subbasal nerves from 1 to 12 months after LASIK. Before LASIK, the subbasal region contained the greatest number of visible nerve fiber bundles. The number decreased in deeper layers and was lowest in the layers destined to become the stromal bed. 

After LASIK, the number of nerve fiber bundles in the subbasal region decreased significantly (P < 0.001) by more than 90%, compared with the number before LASIK. All estimates in this region after LASIK were significantly different from each other, except between 1 week and 1 month and between 1 and 3 months. There was no significant correlation between the intended ablation depth and the change in subbasal nerve fiber bundles from 3 to 12 months after LASIK (r = −0.22, P = 0.39). In the full-thickness stroma and in the stromal flap, the numbers of nerves identified at all times after LASIK were significantly less than estimates before (P < 0.02 and P < 0.001, respectively), and post-LASIK estimates were not significantly different from each other. In the stromal bed there were no significant differences among any estimates before or after LASIK (P = 0.24). 

Discussion

The ophthalmic division of the trigeminal nerve innervates the cornea, principally through the long ciliary nerves. Thick nerve trunks enter the cornea at the limbus and run across the cornea parallel to its surface, often branching several times. Many of these fiber bundles penetrate Bowman’s layer and form the subbasal nerve plexus, with individual fibers terminating in the basal and superficial epithelial cell layers. ^{15 16 17} The central two thirds of the cornea is uniformly and densely innervated, which endows the cornea with an extreme sensitivity to external stimuli. Diminished sensitivity from impaired innervation may reduce reflex tear secretion, reduce avoidance of mechanical or chemical stimuli, and slow the healing response by compromising the epithelium. ¹⁸  

LASIK is considered superior to PRK for correcting moderate and high myopia, because it leaves no corneal haze in the optical axis after surgery. ¹⁹ LASIK damages more corneal nerves than PRK, however, because the LASIK flap is larger than the area of ablation in PRK. Thus, sensory thresholds increase initially. Within the first 6 to 12 months after LASIK, corneal sensitivity returns. ^{3 4 5 6} Our results show, however, that during this recovery the number of nerve fiber bundles that return to the subbasal region, and the stromal flap remains less than half the number before surgery. 

Several investigators have studied reinnervation in histologic sections of corneas after LASIK. ^{7 8 9} Linna et al. ⁷ found that 3 days after LASIK in rabbits, the nerve trunks underlying the wound were preserved in the subjacent stromal bed, and a few regenerative thin nerve fibers were already emerging from the cut stromal nerves at both the flap–bed interface and the wound margin. From 2.5 months after LASIK the regenerated anterior stromal nerves in the flap sent filaments that penetrated the basement membrane and seemed to contribute to the reformation of the basal epithelial and subepithelial nerve plexus. ⁷ In specimens from three human corneas, Latvala et al. ⁸ found no nerves centrally at 8 days after LASIK, but did find subbasal nerves at 54 days and 4 months after LASIK, although they appeared reduced in number. Anderson et al. ⁹ found no nerves in a human corneal specimen 3 months after LASIK and rare, scattered short nerves in a second human specimen 20 months after LASIK. 

Corneal nerves have also been observed in vivo in humans after LASIK by using clinical confocal microscopy, ^{3 12 13 20} although none of these studies attempted to quantify nerve density. In cross-sectional studies, Linna et al. ³ observed subbasal nerve fiber bundles as early as 3 days after surgery in some corneas. Presumably, these nerves either had not degenerated yet or had entered the cornea through the nasal flap hinge that was used. The hinge position may affect postoperative nerve density, because most nerves appear to enter the cornea at the nasal and temporal limbus. ¹⁷ Slowik et al. ¹² did not describe the type of flap in their eight patients examined after LASIK, but they did not observe subbasal nerves in the central cornea for the first 4 months after LASIK. Kauffmann et al. ¹³ examined five eyes longitudinally with a confocal microscope 3, 6, and 12 months after LASIK performed with a nasal hinge and first noted regenerated subepithelial nerve fibers centrally at 6 months. A superior flap hinge was used in the present study, in which at 1 week after LASIK the mean ± SD and median subbasal nerves per scan were 0.3 ± 0.5/mm² and 0.0/mm², respectively, representing nerves in 6 of 17 eyes. ²⁰ It is unlikely that these nerves had regenerated in 1 week, and we assume that they represent nerves that entered the flap through the superior hinge or nerves that had not yet degenerated. 

Oliveira-Soto and Efron attempted to quantify the number of nerves in the central corneas of normal subjects by using confocal microscopy. ¹⁰ They measured 8.7 ± 4.6 nerves per frame in the subbasal plexus compared with our finding of 4.3 ± 1.8 nerves per frame before surgery. Their frames were smaller than ours, however (74,340 μm² vs. 166,250 μm²), so that the differences in nerve density were even greater (117 nerves/mm² vs. 26 nerves/mm² in our study). Three factors may account for this large difference: First, Oliveira-Soto and Efron ¹⁰ counted the number of nerve branches, whereas we counted a branched nerve as a single nerve. Second, Oliveira-Soto and Efron collected continuous images at the corneal depth of interest and selected the “most representative” images for analysis, so that selection bias may have played a role. We used a sampling technique that evaluated all full-thickness scans of sufficient quality to visualize nerves. Third, Oliveira-Soto and Efron used a confocal microscope with more magnification and smaller field size than ours. Perhaps smaller nerve fiber bundles were visualized with their instrument. We know of no studies that compared their instrument with ours. Oliveira-Soto and Efron also measured nerve density as the total length of nerve fibers within a frame. This measurement appears to be more precise and comparable across investigations and could be expressed as nerve fiber length per square millimeter. We hope to use this metric in future studies in conjunction with our sampling technique. 

Nerve fiber bundles are easy to identify by confocal microscopy. They usually appear as bright, well-defined, linear structures. They are sometimes branched, and they usually appear in several consecutive frames. Individual nerve fibers cannot be seen with the clinical confocal microscope, however, and the nerve densities reported herein cannot be directly translated to histologic studies of neurophysiology. Chiou et al. ²¹ identified many linear features by clinical confocal microscopy, including vessels, lattice dystrophy, posterior polymorphous dystrophy, and fungal keratitis, although the only linear structures in normal corneas consisted of nerves. They were well delineated and had homogeneous hyperreflectivity. 

Reinnervation of the cornea after LASIK is similar to the recovery of keratocytes after LASIK. Mitooka et al. ²⁰ noted that at 12 months after LASIK, keratocyte density remained lower than preoperative density in the flap and the anterior retroablation layer, whereas density in the posterior retroablation layer and the posterior stroma did not change after LASIK. Müller et al. ²² demonstrated the direct innervation of individual keratocytes, whereas Vesaluoma et al. ²³ suggested that denervation of the flap may play a role in diminishing the density of keratocytes in this region. The absence of a trophic factor that would normally be supplied by nerves could affect the most anterior keratocytes and decrease their density after LASIK, although it is not clear how much of a role innervation plays in maintaining the health of these cells. Loss of innervation may be only one of several reasons for decreased keratocyte density. 

In summary, we devised a noninvasive method to measure the subbasal nerve density in the central human cornea. Using this method, we demonstrated an immediate decrease in the number of subbasal and stromal nerve fiber bundles in the corneal flap after LASIK. During the first year after LASIK, subbasal and stromal nerve fiber bundles returned to the flap, although by 1 year, their number remained less than half the numbers before LASIK. This loss of innervation may contribute to the loss of keratocytes after LASIK. 

 Supported in part by National Eye Institute Grant EY 02037 (WMB), Research to Prevent Blindness, New York, New York; and Mayo Foundation, Rochester, Minnesota.

 Submitted for publication April 8, 2002; revised June 17, 2002; accepted July 9, 2002.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Confocal images of corneal nerve fiber bundles. (A) Pre-LASIK subbasal nerve fiber bundles; (B) 12-month post-LASIK subbasal nerve fiber bundles (same eye as in A); (C) pre-LASIK nerve fiber bundle (arrow) in anterior stroma destined to be part of the flap; (D) 12-month post-LASIK nerve fiber bundle (arrow) in flap (same eye as in C). Nerves in postoperative corneas often appeared thinner than those before surgery. For printing purposes, total image intensity was adjusted and a median filter applied. Bar, 50 μm.

Figure 2.

 

Stromal layers in the central cornea that were compared before and after LASIK. The layer in the pre-LASIK cornea destined to be ablated (Ablated Zone) was determined from the thickness of the stromal flap and the stromal bed in the 1-month post-LASIK cornea.

Table 1.

 

Number of Nerve Fiber Bundles per Scan per Cornea (n = 17)

	Pre-LASIK	1 Week	1 Month	3 Months	6 Months	12 Months	P *
Subbasal	4.8 (3.5, 5.4)	0.0 (0.0, 0.3)	0.0 (0.0, 0.8)	0.3 (0.0, 1.3)	0.8 (0.4, 1.5)	2.0 (1.0, 3.5)	<0.001, †
	[4.3 ± 1.8]
Stroma	1.0, ‡ (0.6, 1.4)	0.7 (0.4, 1.0)	0.3 (0.0, 0.5)	0.4 (0.0, 1.0)	0.6 (0.4, 0.8)	0.6 (0.3, 0.8)	0.02, §
	[1.1 ± 0.6]
Stromal flap	0.7 (0.5, 1.0)	0.2 (0.0, 0.25)	0.0 (0.0, 0.0)	0.0 (0.0, 0.4)	0.2 (0.0, 0.4)	0.3 (0.0, 0.3)	<0.001, ∥
	[0.8 ± 0.5]
Stromal bed	0.2 (0.0, 0.5)	0.3 (0.2, 0.8)	0.3 (0.0, 0.5)	0.2 (0.0, 0.8)	0.4 (0.0, 0.5)	0.3 (0.2, 0.8)	0.24

 Data are listed as median and interquartile range (Q25, Q75). The means and standard deviations are listed in brackets for the preoperative corneas, which had normally distributed data except for the stromal bed.

*  Friedman test adjusted for multiple comparisons by using the Student-Newman-Keuls procedure.

†  Medians at all times were significantly different from each other except at 1 week versus 1 month and 1 month versus 3 months.

‡  Excludes stromal layer subsequently ablated. If this layer is included (total preoperative stroma), then the values are 1.6 (1.2, 2.0) [1.4 ± 0.5].

§  Median at 1 month was significantly different from median pre-LASIK. All post-LASIK medians were not significantly different from each other.

∥  At all times post-LASIK, medians were significantly different from medians pre-LASIK. Post-LASIK medians were not significantly different from each other.

Figure 3.

 

Median number of central corneal nerves per frame in 17 eyes before and after LASIK. For the subbasal nerves, medians at all times were significantly different from each other (P < 0.001) except at 1 week vs. 1 month and 1 month vs. 3 months. For the stromal flap, medians after LASIK were significantly different from medians before LASIK (P < 0.001), but post-LASIK medians were not significantly different from each other. For the stromal bed, there was no significant change in the median number of nerve fiber bundles (see Table 1 ).