The Effect of Nerve Growth Factor on Corneal Sensitivity After LaserIn Situ Keratomileusis FREE

Laser in situ keratomileusis (LASIK), which is widely used for correctingrefractive errors, combines lamellar corneal surgery with excimer laser photoablation.The automated microkeratome is used to create either a superior or a nasalhinged keratectomy that inevitably truncates a portion of the corneal nervesupply. The nerves of the stromal bed are subsequently exposed to excimerlaser photoablative decomposition. Damaging the corneal nerve supply duringLASIK contributes to a reduction in corneal sensitivity.¹ Theextent and duration of sensory loss depend on the location and number of nervesdamaged by the surgery. Experimental and clinical studies^{2 - 3} havedemonstrated that impairment of corneal sensitivity leads to decreased vitalityand metabolism of the corneal epithelium, frequently associated with epithelialerosion and delay or absence of spontaneous wound-healing capability. Consequently,a rapid recovery of corneal reinnervation is important for restoring the normalphysiological features, tear secretion, and healing properties of the cornea.

Nerve growth factor (NGF) is a polypeptide discovered in the early 1950sby Levi-Montalcini.⁴ It is essential for regulatingthe growth and survival of developing peripheral and central nervous systemneurons. It also serves as a survival and differentiating factor for neuralcrest sensory and sympathetic neurons.⁴ Nervegrowth factor induces neurite sprouting by neuronal cells and restores thefunction of injured neurons.^{4 - 5} Varioustissues and cell types produce and release NGF. Specific high- and low-affinityNGF receptors have been identified on cell membranes, including the ocularsurface.^{6 - 7} Bonini et al⁸ have demonstrated that topical NGF eyedrops improvedcorneal sensitivity and promoted corneal epithelial healing in patients withmoderate and severe neurotrophic keratitis. Although performed in an uncontrolledand nonrandomized series of patients, therapy with topical NGF demonstratedpromise for the restoration of ocular surface integrity and visual functionin those with neurotrophic corneal disease.

While strong evidence suggests that improved corneal sensitivity couldbe attributed to NGF, which induces sensory neuron sprouting into areas ofdenervated cornea,^{8 - 10} nostudy, to our knowledge, has investigated the role of NGF in promoting cornealnerve regeneration after LASIK.

This study investigates whether topically administered NGF plays a rolein accelerating the recovery of corneal sensitivity after LASIK in a rabbitmodel.

Sixteen corneas of 16 New Zealand white rabbits (approximate weight,3.5-4.5 kg) underwent LASIK. All animals were treated according to establishedinstitutional guidelines regarding animal experimentation and the Associationfor Research in Vision and Ophthalmology Regulations for the Use of Animalsin Research.

Intramuscular ketamine hydrochloride (Ketaject), 30 mg/kg of body weight,and intramuscular xylazine hydrochloride (Xyla-ject), 5 mg/kg of body weight,were used to induce anesthesia. A Barraquer-style speculum was placed betweenthe eyelids, and the eye was rinsed with balanced salt solution (BSS). A pararadiallinear mark with gentian violet pencil was applied to the corneal surface.After placement of the suction ring, the intraocular pressure was verifiedto be higher than 65 mm Hg, using a Barraquer tonometer. A nasal-based 160-mm-thickand 8.5-mm-wide hinged corneal flap was created, using an automated microkeratome(SKBM microkeratome; Alcon-Summit Technologies, Cork, Ireland). Subsequently,the microkeratome and the suction ring were removed from the eye and the cornealflap was lifted and retracted against the peripheral cornea.

Excimer laser photoablation was performed on the stromal bed, usingan excimer laser (Summit Apex Plus excimer laser; Summit Technologies, Cork).A single-zone approach (a laser zone diameter of 6.0 mm) was used in all eyesundergoing LASIK. A myopic correction of −3.0 diopters was performedin all eyes for an approximate ablation depth of 36 mm. After the photoablation,the corneal flap was carefully repositioned. A temporary tarsorrhaphy wasthen performed using a 6-0 black silk suture to keep the eyelids closed forthe first week. Antibiotic (0.3% ofloxacin [0.3% Ocuflox; Allergan Inc, Irvine,Calif]) and corticosteroid (0.1% fluorometholone [0.1% FML; Allergan Inc])eyedrops were instilled 4 times a day for the first 7 days.

The treatment agent consisted of murine NGF (200 mg in 1 mL of BSS)purified from the submaxillary gland, as previously described.¹¹ Therabbits received 10 µL of NGF or BSS, using a micropipette. The firsttopical application was applied directly to the exposed stromal bed immediatelyafter photoablation and then onto the corneal surface at a frequency of 4times a day for 3 days. The investigators (M.-J.J.) were masked to the useof NGF solution and BSS.

A prospective study comparing post-LASIK corneal sensitivities of rabbitstreated with topical NGF vs BSS control was performed. A preoperative basicocular examination was performed using a portable slitlamp. The corneal sensitivitywas assessed with an esthesiometer (Cochet-Bonnet esthesiometer; Luneau Ophthalmologie,Chartres, France).¹² Under direct visual control,the nylon filament of the esthesiometer touched the center of the cornea smoothlyand perpendicularly. The diameter of the nylon filament was 0.12 mm, and itslength could be varied from 0 to 60 mm. The pressure applied to the cornea,thus, ranged from 0.4 to 15.9 g/mm². Contact was detected by theslightest bend of the nylon; sensitivity was taken as the length of the filament(in centimeters) that gave a 50% positive corneal reflex (blinking reflex)response from a minimum of 6 stimulus applications. Corneal sensitivitieswere checked twice a week, and the mean of 2 measured sensitivities was usedas the corneal sensitivity of the week.

A t test was used to analyze the differencein corneal sensitivity between treatment with topical NGF and treatment withBSS at given points. Estimates of the average effect of treatments on thecorneal sensitivity over time were calculated using a linear mixed model.¹³ Variability between samples at different points wastreated as a random effect, and variability due to treatment as a fixed effect.The analysis was performed using the MIXED procedure in SAS statistical software.¹⁴

The mean ± SD preoperative measured corneal sensitivity was 4.06± 0.12 cm. The corneal sensitivities of eyes treated with NGF and eyesin the BSS control group were significantly decreased at 2 weeks from preoperativelevels and gradually increased after 2 weeks. The mean corneal sensitivitiesof the NGF-treated group and the control group at every point are shown inTable 1 and depicted inFigure 1. Eyes treated with topical NGF demonstrated an earlierrecovery in corneal sensitivity after LASIK (P =.007) (Figure 1 and Figure 2). A statistically significant difference was found betweeneyes treated with topical NGF and those treated with topical BSS at the post-LASIKpoints of 2, 3, and 4 weeks (Table 1).

The cornea is innervated by sensory nerve fibers originating from theophthalmic division of the trigeminal nerve (cranial nerve V), whose cellbodies lie in the trigeminal ganglion, and by sympathetic nerves, whose cellbodies lie in the superior cervical ganglion.¹⁵ Thenerve fibers penetrate the limbus and form thick nerve bundles in the anteriorthird of the stroma. As these nerves course to the center of the cornea, theybranch horizontally and vertically to form the basal epithelial/subepithelialnerve plexus between the basal epithelial cells and the Bowman layer.¹⁶ Normal corneal sensitivity is essential for propermaintenance of a healthy ocular surface. Experimental and clinical studieshave demonstrated that corneal nerve damage alters the metabolism and vitalityof the epithelium, impairs epithelial healing, and is responsible for trophiculceration.

During LASIK, a corneal flap is created using a microkeratome. Duringthis procedure, the superficial stromal nerves are severed in the lamellarflap margin, and the nerves in the stromal bed under the flap are subsequentlyexposed to laser photoablation. Corneal sensitivity decreases after LASIKbecause of surgical amputation and excimer laser ablation of the nerve fibersinnervating the central corneal surface.^{17 - 19}

Corneal hypoesthesia may consequently trigger a cascade of events thatdegrade the corneal integrity by reducing the protective blinking reflex,delaying epithelial wound healing, and decreasing aqueous tear layer production.^{16 ,20 - 21} Moreover, the lossof corneal sensitivity may disrupt the release of nerve-derived trophic factorsthat are required to maintain the integrity of the corneal epithelium.²² Ultimately, the ocular surface is compromised bychanges resulting in corneal punctate epitheliopathy and corneal surface irregularitythat manifest clinically as ocular surface irritative symptoms and fluctuatingvision.²³

Toda et al²⁴ reported that patients undergoingLASIK developed dry eye with compromised tear function for at least 1 monthafter surgery. Use of artificial tears in the early postoperative period mayhelp ameliorate irritative symptoms and ocular surface damage. Wilson²⁵ reported that staining of the epithelium with rosebengal solution in patients without preexisting dry eye after LASIK is likelya result of neurotrophic epitheliopathy, noting that there is no differencein mean tear production between patients who have significant punctate epithelialerosions and development of rose bengal staining on the flap and those whodo not. The signs and symptoms of LASIK-induced neurotrophic epitheliopathytend to resolve spontaneously approximately 6 months after surgery. This correlateswell with the point when reinnervation of LASIK flaps is complete. Linna etal²⁶ have demonstrated in the rabbit corneathat, at 2.5 and 5 months, more regenerating nerve leashes were observed toemerge from the cut stromal nerve trunks. They seemed to send anastomosingfibers among the neighboring stromal nerves. By this time, the epithelial,basal epithelial/subepithelial, and anterior stromal innervation had gainedan almost normal nerve density and architecture.

In this study, an esthesiometer was used to measure corneal sensitivity.This instrument requires the subject's cooperation not to move during themeasurement, so some skills were needed to obtain reliable data from rabbits,which cannot be trained to cooperate. The tester (M.-J.J. or K.R.Y.) gentlyheld the body of the rabbit and was cautious not to touch the cilia or eyelidswith the instrument. A false-positive reaction was monitored intermittentlyby simply approaching the instrument without touching the corneal surface.Because corneal sensitivity has a topographical variation, data were takenat the apex of the cornea under direct visual control.

Our results suggest that topically applied NGF accelerates the recoveryof corneal sensitivity. We hypothesized that this earlier recovery may leadto an accelerated return of tear function and improvement in the ocular surface.Moreover, the early recovery of corneal sensation may ameliorate the symptomsof dry eyes and associated fluctuating vision after LASIK by restoring tearsecretion from the lacrimal gland and the normal blink reflex.

There have been reports^{8 - 9} thattopically applied exogenous NGF restored corneal integrity and increased thebest-corrected visual acuity progressively during treatment and follow-upin patients with corneal neurotrophic ulcers. The mechanism of action of NGFon the ocular surface in not well understood. Nerve growth factor treatmentmay restore a deficit of synthesis or a release of endogenous NGF. This hypothesisis supported by evidence from an animal study by Brewster et al.²⁷ Inthis study, diabetic rats show reduced expression of target-derived NGF andreduced expression of neuronal genes that are responsive to NGF. The latteris corrected by administration of exogenous NGF. Thus, insufficient neurotrophicsupport might contribute to the pathogenesis of diabetic neuropathy, and anysuccessful treatment might include exogenous neurotrophins or other strategiesto correct their deficiency of action.²⁷ Neurotrophicfactors are proteins that promote the survival of specific neuronal populations.Many have other physiological effects on neurons, such as inducing morphologicaldifferentiation, enhancing nerve regeneration, stimulating neurotransmitterexpression, and otherwise altering the physiological characteristics of neurons.These properties suggest that neurotrophic factors are highly promising aspotential therapeutic agents for neurological disease.²⁸ Anotherindirect mechanism could also be involved, such as increasing neuropeptidesynthesis that promotes epithelial healing or invoking immune cells throughthe release of cytokines.²⁹

The results of this study suggest that topical NGF can induce an earlierrecovery in corneal sensitivity after LASIK. Topical NGF application may beeffective for the early recovery of post-LASIK hypoesthesia. Earlier recoveryshould contribute to a reduction of the symptoms of sensory nerve denervationafter LASIK. A better understanding of the mechanism of action of NGF on theocular surface may allow the broadening of its indications to other externaleye diseases. Further studies to assess the potential of topically administeredNGF to avoid the post-LASIK nerve denervation–induced symptoms are warranted.

Correspondence: Terrence P. O'Brien, MD, Refractive Eye Surgery Service,Ocular Microbiology and Immunology Laboratory, Wilmer Eye Institute, The JohnsHopkins University School of Medicine, 600 N Wolfe St, Woods Building, Room259, Baltimore, MD 21287-9121 (tobrien@jhmi.edu).

Submitted for publication January 10, 2003; final revision receivedJanuary 29, 2004; accepted March 4, 2004.