Experimental Eye Research 200 (2020) 108243

Contents lists available at ScienceDirect

Experimental Eye Research

journal homepage: www.elsevier.com/locate/yexer

Corneal nerves anatomy, function, injury and regeneration

Carla S. Medeiros a, *, Marcony R. Santhiago b, c
a
Santa Casa do Rio de Janeiro, Rio de Janeiro, RJ, Brazil
b
University of São Paulo, São Paulo, SP, Brazil
c
University of Southern California, Los Angeles, CA, United States

A R T I C L E I N F O A B S T R A C T

Keywords: The cornea is a highly innervated tissue, exhibiting a complex nerve architecture, distribution, and structural
Corneal nerves organization. Significant contributions over the years have allowed us to come to the current understanding
Anatomy about the corneal nerves. Mechanical or chemical trauma, infections, surgical wounds, ocular or systemic
Nerve function
comorbidities, can induce corneal neuroplastic changes. Consequently, a cascade of events involving the corneal
Injury
Regeneration
wound healing, trophic functions, neural circuits, and the lacrimal products may interfere in the corneal ho­
Refractive surgery meostasis. Nerve physiology drew the attention of investigators due to the popularization of modern laser
refractive surgery and the perception of the destructive potential of the excimer laser to the corneal nerve
population. Nerve fiber loss can lead to symptoms that may impact the patient’s quality of life, and impair the
best-corrected vision, leading to patient and physician dissatisfaction. Therefore, there is a need to better un­
derstand preoperative signs of corneal nerve dysfunction, the postoperative mechanisms of nerve degeneration
and recovery, aiming to achieve the most efficient way of treating nerve disorders related to diseases and
refractive surgery.

1. Introduction extraction of the lenticule through a small incision to the corneal

surface.
The cornea is one of the most densely innervated tissues in the The subject has reached attention due to the popularization of
human body (Muller et al., 2003). Most of the corneal nerve fibers are of modern laser refractive surgery and the observation of corneal neural
sensory origin, derived from the trigeminal nerve (Muller et al., 1996). structural damage following these procedures. Although nerve damage
In addition to their important sensory functions, corneal nerves help to is inherent to these techniques, it may lead to acute or chronic corneal
maintain the functional integrity of the ocular surface and are respon­ nervous dysfunction that causes patient and physician dissatisfaction
sible for sensations of touch, pain, and temperature. They have an and frustration. This review highlights critical points for identifying and
essential role in blink reflex, wound healing response, and tear pro­ approaching nerve deficiency related to diseases or secondary to corneal
duction and secretion (Buerman and Schimmelpfenning, 1980). They surgery. The comprehension of the nerve injury and regeneration
protect the corneal surface and are essential for epithelial homeostasis. mechanisms after corneal laser surgeries is necessary to properly
Refractive surgery comprises a range of surgical procedures that aim manage the corneal neuropathies and possible factors involved in the
to achieve emmetropia or improved reading in the eye. From the 90s, a neuroregeneration process following refractive surgery.
new era began, with the advent of the excimer laser. Photorefractive
Keratectomy (PRK) was the pioneer among the techniques, correcting
refractive errors on the corneal surface. Later, lamellar techniques were 1.1. Origin, anatomy and nerve organization
implemented, Laser in-situ keratomileusis (LASIK) is the most popular
technique and small incision lenticule extraction (SMILE), a more recent A generous amount of corneal nerve fibers are sensory and originate
procedure. LASIK consists of cutting a flap and ablating the underlying from the ophthalmic branch of the trigeminal nerve (cranial nerve V)
stroma. In turn, the SMILE technique is based on creation of an intra­ (Muller et al., 2003). All mammalian corneas receive sympathetic
stromal lenticule with dimensions dependent on the refractive error and innervation from the superior cervical ganglion (Muller et al., 2003;
Marfurt and Ellis, 1993; Marfurt et al., 1989). Although the presence of

* Corresponding author. Santa Casa da Misericórdia do Rio de Janeiro, Rua Santa Luzia 206, Rio de Janeiro, RJ, Brazil.
E-mail address: [email protected] (C.S. Medeiros).

https://doi.org/10.1016/j.exer.2020.108243
Received 25 June 2020; Received in revised form 6 September 2020; Accepted 8 September 2020
Available online 12 September 2020
0014-4835/© 2020 Elsevier Ltd. All rights reserved.
C.S. Medeiros and M.R. Santhiago Experimental Eye Research 200 (2020) 108243

parasympathetic innervation derived from the ciliary ganglion was re­ reduced function of corneal limbal steam-cells or corneal perforations
ported in cats and rats, their presence in other species is uncertain, (Shannem et al., 2014).
including humans (Muller et al., 2003; Marfurt et al., 1998). Neurotrophic factors are involved in epithelium modulation, prolif­
The nerves enter the cornea at the mid stroma, about 1 mm from the eration, and neuronal growth. Nerve growth factor (NGF) promotes the
limbus, lose their perineurium, and myelin sheath at a mean depth of integrity of the epithelium surface, and when its function is compro­
293 +- 106 μm from the corneal surface. Nerve bundles run horizontally mised there can be the development of neurotrophic ulceration. Brain-
towards the stroma, dividing into progressively smaller branches derived neurotrophic factor (BDNF) enhances epithelial structural for­
forming a stromal network (Muller et al., 1997; Marfurt et al., 2010; mation (You et al., 2000). NGF and BDNF are involved in neuronal
Schimmelpfennig and Beuerman, 1982). growth and regeneration (Namavari et al., 2012). The corneal nerves
Reports differ regarding the number of nerve trunks that infiltrate have a regulatory function of corneal trophic factors, that can be
the stroma through the limbus area, varying from 32.2 to 71.3 nerve compromised after nerve injury.
bundles penetrating the cornea, according to different studies (Muller
et al., 1997; Marfurt et al., 2010). The majority of stromal nerves are 2.1. Clinical features of corneal nerve damage
located in the anterior third of the stroma. The thick nerve trunks are
located bellow the anterior third because of collagen disposition of the Symptoms such as dryness, eye discomfort, pain, fatigue, foreign
stroma, and embedded in an electron-dense extracellular matrix. body sensation, tearing, photophobia, burning, and/or itching may
Important cellular structures as the keratocytes can wrap around adja­ indicate a corneal nerve disorder (Hovanesian et al., 2001; Albeitz et al.,
cent nerve fibers in cytoplasmic extensions, indicating that 2004; Galor et al., 2017). Early identification of factors that may be
cellular-nerve interaction may occur in the cornea (Muller et al., 2003). associated with a nerve dysfunction is crucial to avoid post-surgical
At a certain point, stromal nerves turn 90 degrees and run into the complications related to nerve disorders. A detailed clinical examina­
epithelium, penetrate the Bowman’s membrane, and branch succes­ tion preoperatively is necessary to identify subclinical cases, stabilize
sively into several smaller fibers. Each nerve bundle turns a 90 degree the ocular surface, or even avoid high-risk patients (Shehadeh-Mashor
once more and continues between Bowman’s layer and the epithelial et al., 2019).
basement membrane, parallel to the epithelial layer. At this moment, an One way to monitor the occurrence of nerve-related symptoms is
important structure is organized, the sub-basal plexus (Muller et al., through the clinical use of questionnaires. Validated questionnaires are
2003). The nerves that make up this structure appear beaded, with a available to detect suspected patients with DED or ocular pain (Galor
smooth caliber (Muller et al., 1996; Marfurt et al., 2010; He et al., 2010). et al., 2019). Several methods can be used to study the corneal function.
From the sub-basal plexus, distal epithelial axons originate up to that Esthesiometry is a direct test to evaluates corneal sensitivity perfor­
point of origin. It can give rise to intraepithelial terminals, which can mance, whereas lacrimal function and epithelium vitality are measured
divide into several collateral branches or finish their path as free nerve by indirect tests (Shaheen et al., 2014). The following tests are available
endings in any layer of the epithelium. The intraepithelial terminals may to evaluate tear function changes: tear film breakup time (TBUT), tear
differ in complexity in terms of fiber morphology, orientation, direction, osmolarity, tear volume, and Schirmer tests. Eyelid and meibomian
and length (Muller et al., 1997). According to different studies, gland inspections are essential to detect meibomian gland dysfunction
approximately 351 to 5400–7200 nerve bundles are present in the (MGD), which may lead to evaporative DED. A better analysis of the
human sub-basal plexus, and around 7000 nociceptors mm2 (Marfurt ocular surface health is possible due to the use of corneal dyes, like rose
et al., 2010). The numerical discrepancies found in different studies can bengal, lissamine green, and fluorescein (Wolffsohn et al., 2017).
be attributed to the use of different imaging techniques for the corneal Fluorescein staining happens whenever there is the disruption of a
nerve study. cell-cell junction, allowing the dye diffusion to the intercellular space or
The sub-basal plexus fibers are mostly C fibers (Muller et al., 1996). the stroma (Feenstra and Tseng, 1992). Lissamine green stains ocular
Straight Aδ fibers within the basal epithelial cell layer, run parallel to the surface epithelial cells unprotected by mucin or glycocalyx (Hamrah
cornea surface, but only beaded unmyelinated C fibers travel along the et al., 2011). Rose bengal also stain the cornea and conjunctiva where
sub-basal plexus before turning up perpendicularly and terminate their there is a break of the mucin layer, additionally, stains proliferating
path as free nerve ending (Muller et al., 1997). The cornea contains corneal epithelial cells and affect their viability, lissamine green does
different types of nociceptors, with 70% of them being polymodal and not (Hamrah et al., 2011).
classified as slow-conducting C-fibers, which respond to noxious me­ In vivo confocal microscope (IVCM) allows a detailed corneal nerve
chanical, chemical, and thermal stimuli endogenous inflammatory me­ morphology assessment through a non-invasive method that clinically
diators. Aδ mechanoreceptors are 20% of the receptors and classified as correlates to slit-lamp microscopy and postoperative findings (Shaheen
fast-conducing, transmitting acute pain stimuli, and 10%, Aδ, and C-fi­ et al., 2014). This instrument is capable of studying corneal cellular and
bers are cold thermoreceptors, which are temperature sensitive (Bel­ nerve changes.
monte et al., 1997, 2004). Corneal nerve dysfunctions diagnosis can be challenging since there
is not necessarily a correspondence between signs and symptoms. Thus,
2. Importance of functional integrity of the nerves patients with corneal hypoesthesia can be asymptomatic, although
exhibiting severe ocular surface involvement. On the other hand, pa­
Corneal sensory nerves carry nerve stimulus to the brain via afferent tients with corneal allodynia may not have significant changes in the eye
nerves, where they are interpreted and return through efferent stimuli, exam but present severe discomfort (Belmonte et al., 2004; Shaheen
passing within an autonomic route to the lacrimal glands. After neuronal et al., 2014).
damage, the neural circuit is somehow interrupted. Corneal infections,
trigeminal nerve damage caused by surgery, trauma, or neurological 2.2. Clinical nerve disorders related to refractive surgery
disease, can interfere with this circuit’s function. Consequently, it cre­
ates changes in the blink reflex, tears function, corneal sensitivity. Thus, 2.2.1. Dry eye after refractive surgery
affecting the tissue homeostasis and generating signs and symptoms of Dry eye disease (DED) is multifactorial and characterized by loss of
corneal neuropathies. homeostasis, in which tear film instability and hyperosmolarity, ocular
The corneal innervation plays a significant trophic influence on the surface inflammation, and neurosensory abnormalities have etiological
corneal epithelium, promoting a healthy ocular surface. Thus, a corneal roles (Wolffsohn et al., 2017). DED can lead to impaired best-corrected
nerve deficiency can be exhibited through clinical signs as persistent vision due to tear film instability, symptoms such as visual fluctuation,
epithelial defects (PED), neurotrophic keratitis, dry eye disease (DED), a glare, worsening night vision, and problems related to computer vision

2
C.S. Medeiros and M.R. Santhiago Experimental Eye Research 200 (2020) 108243

(Shehadeh-Mashor et al., 2019). Thus, explaining why this ocular 2007; Ambrosio et al., 2008; Shehadeh-Mashor et al., 2019).
diagnosis may represent one of the biggest causes of patient dissatis­ Also, the resulting deficiency of trophic factors that regulate the
faction after refractive surgery (Levinson et al., 2008). metabolism and viability of corneal cells is associated with decreased
Risk factors for the development of DED after refractive surgery are sensitivity and altered healing, leading to a neurotrophic state of the
older patients, women, contact lenses users, and the occurrence of DED cornea (Ambrosio et al., 2008). The lacrimal gland production of cyto­
symptoms before surgery (Shehadeh-Mashor et al., 2019; Albeitz et al., kines increases in response to an epithelial wound. These observations
2004). A lower preoperative refractive error was suggested to be a risk suggest there is a nerve-mediated feedback from the cornea to the
factor for DED by Shehadeh-Mashor et al. (2019), whereas greater lacrimal glands that modulates the expression of nerve growth factors
ablation depth and attempted correction was found to correlate with and cytokines responsible for homeostasis and the modulation of
higher incidence of DED by Albietz et al. (2004). epithelial wound healing processes (Ambrosio et al., 2008). Thus, nerve
transection during LASIK changes these corneal functions.
2.2.2. LASIK-induced neurotrophic epitheliopathy (LINE)
Wilson and Ambrosio suggested the occurrence of LINE as a result of 2.2.3. Surgical induced pain
neurotrophic and inflammatory compounds (Fig. 1) (Wilson and Postoperative acute pain remains the main limiting factor of PRK for
Ambrósio, 2001). The disease can manifest itself through a combination laser vision correction. It is a variable and unpredicted event that can be
of signs and symptoms, such as punctate epithelial erosions detected severe in a significant amount of the patients (Garcia et al., 2016). The
through lissamine green or rose bengal along the flap, associated with peak of these symptoms is at 48 hours and usually declines 72 hours
the best-corrected vision worsening, visual fluctuation and night vision following the surgery with the advance of corneal reepithelization
problems over the first weeks after the procedure (Ambrósio et al., (Cherry et al., 1994). However, some patients may experience sensitivity
2008). modifications and aberrant sensations weeks or months after corneal
The underlying mechanism of LINE is the surgical amputation of surgery due to the persistence of nerve excitability disturbance in the
nerves at the flap edge in addition to laser ablation of nerves causing a damaged or regenerated nerves (Gallar et al., 2007).
significant reduction in corneal sensitivity. Hypoesthesia results in the The degree of eye discomfort caused by LASIK in the early post­
interruption of the trigeminal afferent nerves that course to the brain­ operative period is significantly lower than that of PRK. However, long-
stem, resulting in interruption of the nerve circuit to the lacrimal gland. term symptoms after LASIK can persist as long as two to five years, even
Consequently, this impacts lacrimal gland tears production and the blink after the resolution of objective signs of neuron disorders (Tsuiski et al.,
rate (Ambrosio et al., 2008). The resultant tear film hyperosmolarity 2007). Increasing evidence explains manifestations of ocular discomfort,
drives the accumulation of inflammatory mediators resulting in an in­ commonly known as “dry eye,” is better understood as corneal pain,
flammatory state of the corneal surface, which is also induced by the especially after LASIK, suggesting that ocular pain following LASIK in­
femtosecond and excimer laser (Sambursky and O’Brien, 2011). Other dicates a disordered hypersensitivity of the ocular somatosensory nerves
hypotheses regarding the possible mechanisms are the corneal shape injured after surgery. About 20–50% of the patients will develop at least
modification following laser ablation, affecting the relationship between mild symptoms of dry eye/persistent neuropathic pain after LASIK
the ocular surface and eyelids, resulting in an abnormal tear film dis­ (Levitt et al., 2015).
tribution, and increased evaporation (Lee et al., 2000). Plus, a hypoth­ A variety of symptoms can encompass an altered perception of pain,
esis regarding the suction devices used to create flaps on for example, pain evoked by wind (hyperalgesia), an increased sensi­
microkeratomes and femtosecond lasers, producing limbal globet cell tivity beyond the injured area (secondary hyperalgesia), dysesthesia
damage, and consequently a reduction on globet cell mucin leading to (burning), or pain due to harmless stimuli (allodynia) (Levitt et al.,
tear film instability (Wilson and Ambrosio, 2001; Rodriguez-Prats et al., 2015). The persistence of pain without objective signs in the ocular
exam is a pattern suggestive of neuropathic pain (Vehof et al., 2017;
Galor, 2019). Neuropathic pain, corneal allodynia, and hyperalgesia are
clinical counterparts. The pain perception in the cornea is due to the
transduction of a noxious stimulus above the acceptable level, creating a
hyperexcitability status and triggering an electrical peak in nociceptors
located at the terminals of C and Aδ fibers on cornea (Rosenthal et al.,
2009).

2.2.4. Recurrent corneal erosion (RCE) and persistent epithelium defects

(PED) after refractive surgery
The incidence of RCE is 0.05%–18% (Seiler, 1993; Diez-Feijóo et al.,
2014), but those numbers could be underreported due to the absence of
clinical signs in mild cases (Hovanesian et al., 2001). Acute pain that
wakes the patient, lacrimation, photophobia, foreign body sensation,
soreness of the eyelid to touch, eyelid sticking, and sharp pain may
indicate sub-clinical RCE possibly happening even long-term after
refractive surgery (Hovanesian et al., 2001; Lin et al., 2019). Typical
IVCM findings are epithelial microcysts and an epithelial basement
membrane (EBM) that protrudes into the epithelium or the damaged
sub-basal plexus (Miller et al., 2019).
The refractive surgeon needs to pay attention to the subclinical signs
of the epithelial basement membrane corneal dystrophy (EBMD). PRK
was pointed out as a risk factor for RCE (Diez-Feijóo et al., 2014),
whereas LASIK is considered a contraindication in eyes with EBMD by
some surgeons (Miller et al., 2019). However, recurrent erosion
Fig. 1. Lasik-induced neurotrophic epitheliopathy (LINE). following PRK seems paradoxical, since Phototherapeutic Keratectomy
The image shows ocular surface punctate epithelial erosions on the surface of a (PTK) is one of the options of treatment of recurrent erosions (Wilson
LASIK flap. et al., 2017). The RCE occurs in the presence of a dysfunctional or

3
C.S. Medeiros and M.R. Santhiago Experimental Eye Research 200 (2020) 108243

injured EBM, so the ablation of this ineffective structure is necessary to 3. Surgery-induced corneal nerve dysfunction
promote the adherence of the corneal epithelium to the underlying
stroma, so, a new and effective EBM can properly connect a healthy 3.1. Nerve damage following LASIK
epithelium to the underlying stroma. On the other hand, once PRK
creates a new stromal bed, it can also be the triggering cause of RCEs. During LASIK, massive nerve damage occurs, reaching 80–90% of
Thus, the same procedure can be the cause and the treatment, and the nerve decrease compared to preoperative values. Linna et al. (1998)
data available on the current subject cannot explain this paradoxical demonstrated the total absence of basal/epithelial and sup-epithelial
situation (Diez-Feijóo et al., 2014). nerves in the flap 3 days after LASIK, except in the hinge area. The
Another undesirable complication related to PRK is the delay of nerve deficiency can persist to 3 weeks to 5 years after surgery according
epithelium healing, leading to a PED. The persistence of an epithelial to different studies (Chuck et al., 2000; Moilanen et al., 2008; Calvillo
defect that does not heal ten days to two weeks following surgery usually et al., 2004; Erie et al., 2005) (Table 1). A correlation between the loss of
develop corneal scarring of the anterior stroma beneath the defect, even sensation and an abnormal nerve morphology was seen following the
if posteriorly the epithelium heals (Wilson et al., 2018). Within an procedure (Linna et al., 2000). As a result of the surgical lesion, corneal
epithelial defect that persists, there is no regenerated EBM. Conse­ hypoesthesia is present from 3weeks to 12 months (Chuck et al., 2000;
quently, there is a constant influx of pro-fibrotic growth factors that Kauffmann et al., 1997) (Table 1). The corneal sensitivity is first restored
drives myofibroblast differentiation, extracellular matrix disorganiza­ compared to the baseline nerve density (Bragheeth and Dua, 2005) and
tion, and thus, the loss of transparency of the corneal tissue (Jester et al., is higher in the hinge area.
2012). Cutting the flap implies a nerve fiber disruption at the edge of the
wound, which varies in length and depth according to the surgical plan.
A more significant nerve impairment was observed in hyperopic abla­
tions compared to moderate or low myopic treatments. A bigger flap and
deeper ablations (Kim and Kim, 1999) are related to a more significant
corneal sensitivity loss and a delayed reinnervation restoration

Table 1
Corneal nerve function and impairment after LASIK: summary of published studies.
LASIK

AUTHORS N (EYES) CN SENSITIVITY LOSS CN SENSITIVITY RECOVERY CN DENSITY LOSS CN DENSITY

RECOVER

Agca et al. (2015) 15 – – 3 m: 26% of pre-op 6 m: 33.5% of pre-

SBP density loss higher after than op †
SMILE up to 3 m
Bragheeth and Dua 83 1 and 3 m: A Higher decrease in high 3 m: moderate myopia; 6 m: Lower SBP density in high myopia 12 m †
(2005) myopia and hyperopia groups compared high myopia and hyperopia and hyperopia groups than in
to moderate myopia groups. moderate myopia until 6 m
Calvillo et al. (2004) 17 – – 1 m: 10% of pre-op * 3 y: 60% of pre-op
†
Chuck et al. (2000) 28 1 day 3w – –
Darwish et al. (2007) 20 1m* 3m 1 m: 23.3% of pre-op * 6 m: 60.5% of pre-
op †
Demirok et al. (2013) 28 1w * CN sensation loss higher than SMILE 6 months – –
until 3 m
Erie et al. (2005) 16 – – 1 y: 49% of pre-op values * 5y
Ishi et al. (2015) 30 1w: 21.53% of pre-op * 12 m: 31.4% †
Kauffmann et al. 15 4m 12 m 10 w 12 m †
(1997)
Kim and Kim (1999) 40 – 6m† – –
Lee et al. (2006) 56 3 m = 40% of pre-op* 6 m: 59.6% of pre-op † 3 m: 29% of pre-op * 6m†
6 m: 31% of pre-
op
Li et al. (2013) 22 1 w * Lower CN sensitivity than SMILE up 6m† 1w* 6 m † 6 m: 38% of
to 6 m 3 m: 12.15% pre-op
SBP density loss higher than SMILE
up to 3 m
Linna et al. (2000) 56 3 d started but reached the lowest by 1-2w 6 months – –
Mohamed-Noriega 12 – – 1w* 4w: †
et al. (2014)
Moilanen et al. (2008) 15 – – 5 days: 18% of pre-op 2 y: 64% of pre-op
levels. †
Nejima et al. (2005) 115 1 w * Lower CN sensitivity than PRK up to 12 m † – –
3m
Patel et al. (2010) 42 1m* 3m 1m* 36 m † 36 m: 65%
of pre-op values
Reinstein et al. (2015) 1891/21 1d 12 m in 10 of 16 studies – –
studies
Wei and Wang (2013) 54 1w * Lower than SMILE up to 3 m 3m† – –
Xia et al. (2016) 59 1d 6m – –

Legend: N= Number of eyes, CN= Corneal nerves, SBP= Sub-basal plexus, NFB = nerve fiber bundles, pre-op = pre-operative, m = months, w = weeks, d = days, * =
First time-point studied, † = preoperative values have not been reached until the last time-point, y = years.
The columns represent parameters to analyze the loss and recovery of the corneal nervous function, as well as a numerical measurement of the nerve density and
organization, aiming to understand the nerve injury and regeneration after the procedure. The rows represent different studies, as well as their results.

4
C.S. Medeiros and M.R. Santhiago Experimental Eye Research 200 (2020) 108243

(Bragheeth and Dua, 2005). Another variable in the LASIK equation has et al., 1992) (Table 2). Superior hypoesthesia and lower tear break up
been shown to have an impact on the final nerve population, the corneal time were reported after LASIK compared to PRK eyes, although reduced
hinge. A more significant number of fibers were spared in wider hinge tear secretion was observed in both procedures, with no significant
flaps compared to narrow (Donnenfeld et al., 2004), as well as the hinge differences noted (Nejima et al., 2005).
location also seems to have an impact on corneal sensitivity and dry eye Mitomycin C (MMC) is a quinolone antitumor drug used in refractive
occurrence in the early postoperative period that does not persist after surgery prophylactically and therapeutically to regulate the develop­
six months (Feng et al., 2013). ment of corneal haze as a complication of surface ablation surgery. This
potent inhibitor of cell proliferation and myofibroblast generation has
proven to be effective in modulating wound healing and, therefore, is
3.2. The impact of PRK and MMC on corneal nerves
currently used as an adjuvant for treating and preventing corneal haze
(Santhiago et al., 2012). This drug has a known cytotoxic effect, causing
Nerve loss after PRK has several different etiologies. First, during
expressive keratocyte apoptosis on the anterior stroma following PRK.
epithelial scrape, all the delicate intra-epithelial nerve fibers and the
However, MMC is not cell cycle-specific, and a neurotoxic effect on
sub-basal plexus are wounded. Secondly, the nerve damage reaches the
peripheral nerves was reported, including ocular tissues (Kornek et al.,
anterior to medium stroma, depending on the planned level of excimer
1998; Sui et al., 2014; Mietz et al., 1995, 1997). The additional use of
laser ablation. The corneal nerve injury following PRK extends beyond
MMC during PRK caused extra nerve damage on the day following
area affected by the excimer laser ablation through the process
surgery. However, the neurotoxic effect on cornea does not persist after
referred to as retrograde axonal death. Thus, the nerve loss is more
the first month. Probably due to the single use of the drug, restricted to
extensive than the diameter and depth of the laser ablation because of
the intraoperative application, and its efficient elimination from the
the extension of the axonal death (Medeiros et al., 2018) (Fig. 2).
superficial cornea (Medeiros et al., 2018). Therefore, despite the early
The nerve fiber degeneration starts after the procedure and can last 7
neurotoxic effect, the additional use of MMC in PRK does not cause
days. According to different authors, the corneal nerve population re­
long-term nerve-related concerns (Gambato et al., 2011; Medeiros et al.,
mains diminished for up to 12 months after PRK (Table 2). However,
2018).
persistent decreases in nerve function and abnormal morphology of the
sub-basal plexus was noted in some studies up to the last time-point
(Trabucchi et al., 1994; Kauffmann et al., 1997; Medeiros et al., 2018) 3.3. SMILE and nerve loss
(Fig. 5). The morphology and nerve distribution of the corneas that had
PRK is thought by many be comparable to control eyes by five years after Small incision lenticule extraction (SMILE) is a flapless procedure;
surgery. However, in one study, eyes exhibited a persistent reduction in consequently, a minor neural remodeling and better biological healing
nerve density and, therefore, did not achieve complete neural recovery may result in a lower impact on the ocular surface than other procedures
at 5 years after surgery (Moilanem et al., 2003). The corneal sensitivity (Denoyer et al., 2015; Xia et al., 2016; Recchione et al., 2020; Lin et al.,
is impaired after PRK, around 75% of the preoperative levels (Campos 2019). A lower drop in corneal sensitivity was observed after SMILE

Fig. 2. Retrograde Axonal Death.

Beta-III tubulin histochemistry in A) and C) a control
rabbit cornea in the left column, compared to B) and D) a
cornea 1 day after -9.0 D PRK without MMC in the right
column. The first row shows the corneal innervation in
the sub-basal plexus and second row in the mid-stroma.
This shows the initial nerve destruction caused by the
action of the excimer laser. The excimer laser ablation
zone is represented by the white circle. In the first row,
we observed at the sub-basal plexus (50-75μm depth)
that the nerve damage occurred at a wider diameter than
the ablation zone. In the mid-stroma, nerves in the con­
trol eye (C) and after 1 day of PRK (D), respectively, are
at a depth of 225-250μm. Thus, nerve injury extends
beyond the excimer laser ablation. Mag 20X.

5
C.S. Medeiros and M.R. Santhiago Experimental Eye Research 200 (2020) 108243

Table 2
Corneal nerve function and impairment after PRK: summary of the results of published studies.
PRK

AUTHORS N CN SENSITIVITY LOSS CN SENSITIVITY CN DENSITY LOSS CN DENSITY RECOVERY

(EYES) RECOVERY

Campos et al. (1992) 14 1 m: 75.2% of pre-op 3 m: 95.7% of pre-op – –

Erie et al. (2005) 18 – – 1 y: 41% of pre-op 2y
Frueh et al. (1998) 18 – day 0 12 m: 75% of the eyes. † ‡
Gambato et al. 28 – 5y‡
(2011)
Heinz et al. (1997) 25 5–8 m ‡
Kauffmann et al. 15 6–12 m – 6–8 m ‡
(1997)
Linna and Tervo 10 7d 12 m †
(1997)
Medeiros et al. 28 1 d: 61% of pre-op values of 3 m: 95% of pre-op values of nerve
(2018) nerve area area ‡
Moilanem et al. 14 5 y: 71% of eyes
(2003)
Nejima et al. (2005) 28 1w; Higher CN sensitivity than LASIK 6m –
up to 3 m
Tervo et al. (1994) 17 1h 12 m ‡
Trabucchi et al. 34 day 0 4m‡
(1994)

Legend: N= Number of eyes, CN= Corneal nerves, SBP= Sub-basal plexus, NFB = nerve fiber bundles, pre-op = pre-operative, m = months, w = week, d = day, * =
First time-point studied, y = years, † = preoperative values have not been reached until the last time-point, ‡ = morphological changes persisted after the last time-
point.
The columns represent parameters to analyze the loss and recovery of the corneal nervous function, as well as a numerical measurement of the nerve density and
organization, aiming to understand the nerve injury and regeneration after the procedure. The rows represent different studies, as well as their results.

compared to FSLASIK (femtosecond-LASIK) (Xia et al., 2016; Wei et al., SMILE.

2013; Li et al., 2013; Reinstein et al., 2015; Demirok et al., 2013)
(Table 3). During SMILE, a 3–4.5 mm incision length is performed, and a 4. Nerve regeneration after surgical lesion
large portion of the anterior stromal nerves in the cap are preserved (Li
et al., 2013). In contrast, all the nerve fibers from the epithelium to the 4.1. Nerve recovery after LASIK
anterior stroma are transected during LASIK (Li et al., 2013). Recchione
et al. (2020) quantified the damage to the sub-basal plexus after An initial regeneration process starting 3 days after LASIK has been
FSLASIK to be a reduction of 75% of the preoperative values. In described, in wich newly formed subepithelial fibers originate new
contrast with a 23% reduction in the SMILE group (Recchione et al., epithelial nerves (Linna et al., 1998). Sprout neurites penetrate the up­
2020). Increased tear osmolarity, lower tear breakup time (TBUT) and permost stroma and infiltrate the basement membrane, reassembling the
Schirmer were more significant after LASIK compared to SMILE sub-basal plexus around two and a half months after LASIK (Linna et al.,
(Demirok et al., 2013; Ganesh et al., 2014; Xia et al., 2016), 1998). The nerves recovering after LASIK migrate from the periphery to
suggesting better tear film performance and ocular surface health after the center of the wound, in a centripetal growth pattern (Kauffmann

Table 3
Corneal nerve function and impairment after SMILE: summary of the results of published studies.
SMILE

AUTHORS N (EYES) CN SENSITIVITY LOSS CN SENSITIVITY CN DENSITY LOSS CN DENSITY

RECOVERY RECOVERY

Agca et al. (2015) 15 – – 3 m: 36% of pre-op values 6 m: 33.5% of pre-op

SBP density higher than FSLASIK values†
until 3 m
Demirok et al. (2013) 28 1w*; CN sensation loss higher in 6m – –
FSLASIK until 3 m
Ishi et al. (2015) 30 – – 1w: 47.4% of pre-op values 12 m: 70,7% †
Li et al. (2013) 32 1w * Higher CN sensitivity than 6m† 1 w * 3 m: 36.3% of pre-op values 6 m: 45.4% of pre-op
FSLASIK up to 6 m SBP density higher than FSLASIK values †
until 3 m SBP
Liu et al. 2015 30 – – 1w* 3m
Mohamed-Noriega et al. 12 – – 1w * 1m
(2014)
Reinstein et al. (2015) 330/ 1d 6 m (6 of 8 studies) – –
7studies
Wei and Wang (2013) 61 1 w * CN sensitivity higher than 3m – –
FSLASIK up to 3 m
Xia et al. (2016) 69 1d 1m – –

Legend: N= Number of eyes, CN= Corneal nerves, SBP= Sub-basal plexus, NFB = nerve fiber bundles, pre-op = pre-operative, m = months, w = week, d = day, * =
First time-point studied, y = years, † = preoperative values have not been reached until the last time-point.
The columns represent parameters to analyze the loss and recovery of the corneal nervous function, as well as a numerical measurement of the nerve density and
organization, aiming to understand the nerve injury and regeneration after the procedure. The rows represent different studies, as well as their results.

6
C.S. Medeiros and M.R. Santhiago Experimental Eye Research 200 (2020) 108243

et al., 1997; Cavanagh et al., 1993). Medeiros et al. (2018) described in an animal study the nerve
Stromal nerves that underwent transection during the flap cut were regeneration behavior of the corneal nerves following PRK. After the
not restored even five months after LASIK. However, most of the thick first month, an initial attempt to restore the nerve population exhibited
nerve bundles sent thin regenerated fibers, as a sign of previous neural wound-oriented neurites. The primary nerves to repopulate the cornea
damage. Therefore, it is difficult to distinguish if these fibers survived migrated from the sub-basal plexus towards the epithelium. The anterior
after LASIK or are regenerated from Schwann cells (Linna et al., 1998). stroma also revealed neurites, and the new nerves grew perpendicularly
Unlike PRK, Schwann cell channels remain preserved despite the flap to the edge of the wound (Fig. 3). Despite using different surgical lesion
cut. Thus, the cells act as guides to regenerate neurites in neighboring techniques, a similar neuron regeneration behavior was previously
nerve trunks. The integrity of the Bowman’s layer may facilitate the described (Rozsa et al. 1983) This study provided several relevant
infiltration of newly formed stromal fibers that contribute to the contributions about neurogenesis following a corneal lesion.
extension of subepithelial neuritis. Also, the absence of sub-epithelial The mid-stroma showed signs of nerve rehabilitation after the first
scarring seems to facilitate the formation of new epithelial neurites month of PRK. The recovered nerve fibers are marked by the presence of
(Linna et al., 1998). a primitive innervation trying to repair the ablated area (Medeiros et al.,
2018). The nascent nerves revealed distinct morphologic characteristics,
particularly the wound margin, presenting multiple branches and
4.2. Neuroregeneration after PRK increased tortuosity (Fig. 3). These findings mark the first phase of
neuron regeneration after PRK (Medeiros et al., 2018; Rózsa et al.,
The neuroregeneration event following laser injury is initiated at 4–6 1983).
weeks after PRK (Trabucchi et al., 1994; Medeiros et al., 2018). Studies The second month following PRK, there was degeneration of neurites
about the time necessary to restore the nerves differ considerably. The that first began to repopulate the cornea and the absorption of axon
literature demonstrated that the regenerated nerve area reached a vestiges (Fig. 4). Thus, the second phase of nerve regeneration was
comparable value with pre-operative eyes from 3 months to 5 years initiated, characterized by the formation of dense neurites, in an oblique
(Table 2). Many differences regarding the methodology used can explain arrangement, replacing the immature innervation (Medeiros et al.,
the variation among different reports concerning the time to achieve 2018) (Fig. 4). This neuron behavior is considered the first sign of
nerve recovery. IVCM is an essential non-invasive and in-vivo resource. neurogenesis (Rózsa et al., 1983) and can be seen three months after
However, IVCM images are recorded from the corneal apex, which can PRK (Fig. 4). However, the recovered nerves appeared mainly in the
repeatedly record the nerve bundles. Thus, an overlap may occur at the anterior stroma. A substantial population of intraepithelial nerves did
edge of each image. The imaging technology applied is critical when not reach the central corneal area and still an abnormal architecture and
studying the corneal nerves, especially after PRK, due to backscattering orientation persisted. Thus, neural remodeling was not complete by six
light, which can negatively impact the final image quality, and conse­ months (Medeiros et al., 2018) (Fig. 5).
quently, the data (He et al., 2010).

Fig. 3. Neural remodeling one month after PRK.

Beta-III tubulin histochemistry of a control rabbit
cornea in the left column compared to 1 month after
-9.0 D PRK without MMC in the right column. Each
row shows different layers in the cornea: sub-basal
plexus and mid-stroma, respectively. This image
shows an aborted attempt to recover the corneal
innervation. A) Normal sub-basal plexus distribution.
B) At the intraepithelial and sub-basal plexus levels,
the nerves grew perpendicular to the margin of the
surgical lesion (arrows). C) Normal mid-stroma
innervation, and D) the aberrant morphology of
nascent mid-stromal nerves at 1 month after PRK
(arrows). The latter nerves were characterized by
disorganized aspect, with multiple branches,
concentrated at the wound edge. Mag. 20X.

7
C.S. Medeiros and M.R. Santhiago Experimental Eye Research 200 (2020) 108243

Fig. 4. The second phase of nerve regeneration.

Beta-III tubulin histochemistry in a rabbit cornea at
two and 3 months after -9.0 D PRK without MMC. A)
Demonstrates a cornea at 2 months after -9.0 D PRK,
and at this point the first sign of nerve regeneration is
noted in the mid-stroma, with the degeneration of
neurites that reinnervated the cornea at 1 month after
PRK (see Fig. 3D) and the absorption of axonal traces
(arrows). Thus, at two months after PRK, there was an
overlap between the two neuroregeneration phases.
B) A classic sign of the second phase of neuro­
regeneration was present at month 3, with the oblique
arrangement of the intra-epithelial/sub-basal plexus
(arrows). C) In month 3, the anterior stromal nerves
were thinner and more numerous than in the control
unwounded cornea, which indicates previous nerve
degeneration. D) Normal rabbit cornea anterior stro­
mal nerves. Mag. 20X.

Fig. 5. Neural remodeling six months after PRK.

Beta-III tubulin histochemistry of a control rabbit
cornea in the left column, compared to a cornea at 6
months after -9.0 D PRK without MMC in the right
column. A) Shows the normal sub-basal plexus
arrangement. B) Shows a lack of nerve fibers in the
central area of sub-basal plexus at 6 months after
PRK (asterisk). Thus, the neuroregeneration process
was not completed by the sixth month after PRK. Mag
20X.

Recent evidence revealed that regenerating nerves avoid corneal 4.3. Reinnervation after SMILE
regions populated by myofibroblasts. Thus, following PRK, the α-smooth
muscle actin (α-SMA) positive cells inhibited stromal nerves in the ab­ Faster corneal nerve sensitivity and density recovery were reported
lated zone from repopulating the sub-basal and epithelial layers above after SMILE than FSLASIK (Table 3) (Mohamed-Noriega et al., 2014; Xia
them (Jeon et al., 2018). These research findings could explain the et al., 2016; Ganesh et al., 2018). Nevertheless, the morphological
delayed sub-basal and intra-epithelial nerves recovery noted in the ab­ regeneration patterns did not differ between the two techniques (Li
lated area in another study (Medeiros et al., 2018). Another hypothesis et al., 2013). The reinnervation mechanism is centripetal, with the
is that the sub-basal and sub-epithelial network, a complex and dense nerves regenerating from the end of severed stromal fibers peripherally
nerve structure, might need a longer recovery time due to slow neural towards the center, passing the transected nerve cap, aiming to establish
recovery. a new sub-basal plexus (Li et al., 2013). A regular arrangement of the
nerve fibers following LASIK reinforces the hypothesis that a precise

8
C.S. Medeiros and M.R. Santhiago Experimental Eye Research 200 (2020) 108243

realignment of distal Schwann cell channels that are preserved after The topical use of anesthetics, morphine, cycloplegics, steroids,
transection, could facilitate the re-connection and regrowth of damaged contact lens, and NSAIDs may be individually considered and carefully
nerve end at the flap or the cap (Li et al., 2013). administered on corneal pain management (Woreta et al., 2013).
The corneal sensation recovery begins with 4 weeks but reaches However, oral medications may be necessary: oral analgesics, opiates,
preoperative levels at 6 months in the majority of the studies (Reinstein NSAIDs are also treatment options (Garcia et al., 2016). Other systemic
et al., 2015) (Table 3). However, full nerve density regeneration can take drugs, such as gabapentin, pregabalin, or anti-depressants, can be used
more extended periods, up to 12 or 24 months (Ishi et al., 2015; Ves­ as adjucants to treat ocular pain (Woreta et al., 2013).
tergaard et al., 2013) (Table 3). When conservative therapy fails in the treatment of RCE, the surgical
approach is an option. Anterior stromal puncture, epithelial alcohol
5. Corneal nerve wound healing and nerve regeneration delamination, amniotic membrane, and PTK are successful therapies for
pathways RCE (Lin et al., 2019; Wilson et al., 2017).

The correlation between compromised innervation, epithelial ho­ 7. Conclusion

meostasis, and deficient wound healing was previously published
(Buerman and Schimmelpfennig, 1980). An in-vivo study demonstrated Refractive surgery causes significant nerve loss, and can lead to
corneal fibroblasts promoting neurite outgrowth in cultured chick cells nerve-related disorders. It is crucial for the surgeon to recognize the
in a dose-dependent manner (Yam et al., 2017). Later, Jeon et al. (2018) nerve dysfunction, since corresponding signs and symptoms are often
described in vitro, a distinct neuronal behavior between myofibroblasts not present. Thus, a better comprehension of the pathogenesis of nerve
and fibroblasts co-cultures with sensory neurons. While myofibroblasts’ degeneration and regeneration facilitates efficient management, pre­
contact inhibited neurite outgrowth, nerves grew in the typical fashion vents further nerve destruction, or stimulates nerve regeneration.
through fibroblast cell bodies. Myofibroblasts co-cultured with sensory
neurons also caused delayed neurite outgrowth (Jeon et al., 2018). This References
study also made important observations about the wound healing
response and nerve regeneration in cats. Regenerated nerves avoid Agca, A., Cankaya, K.I., Yilmaz, I., Yildirim, Y., Yasa, D., Olcucu, O., Demircan, A.,
corneal regions populated by myofibroblasts in corneas after PRK. Demirok, A., Yilmaz, O.F., 2015. Fellow eye comparison of nerve fiber regeneration
after smile and femtosecond laser-assisted LASIK: a confocal microscopy study.
However, greater nerve regeneration was reported in corneas with the
J. Refract. Surg. 31, 594–598.
additional use of Mitomycin C during PRK, once a smaller zone of Albeitz, J.M., Lenton, L.M., McLennan, S.G., 2004. Chronic dry eye and regression after
α-SMA-positive myofibroblasts was present . Thus, fewer myofibroblasts laser in situ keratomileusis for myopia. J. Cataract Refract. Surg. 30, 675–684.
promoted nerve fiber repopulation (Jeon et al., 2018). Ambrósio Jr., R., Tervo, T., Wilson, S.E., 2008. LASIK-associated dry eye and
neurotrophic epitheliopathy: pathophysiology and strategies for prevention and
Recently, the effect of MMC affecting neurite outgrowth was treatment. J. Refract. Surg. 24, 396–407.
confirmed (Hindman et al., 2019). The author also studied the effect of Belmonte, C., GarciaHirschfeld, J., Gallar, J., 1997. Neurobiology of ocular pain. Prog.
another anti-fibrotic commonly used after PRK surgery, 1% predniso­ Retin. Eye Res. 16, 117–156.
Belmonte, C., Acosta, M.C., Gallar, J., 2004. Neural basis of sensation in intact and
lone acetate. A more persistent α-SMA-positive myofibroblast zone, injured corneas. Exp. Eye Res. 78, 513–525.
peripheral acellular areas, peripheral intra-epithelial nerve reduction, Bragheeth, M.A., Dua, H.S., 2005. Corneal sensation after myopic and hyperopic LASIK:
and central corneal reinnervation delay were more pronounced in the clinical and confocal microscopic study. Br. J. Ophthalmol. 89, 580–585.
Buerman, R.W., Schimmelpfenning, B., 1980. Sensory denervation of the rabbit corneas
PRK plus prednisolone group than in the PRK only or PRK-MMC groups affects epithelial properties. Exp. Neurol. 69, 96–201.
(Hindman et al., 2019). In this study, Hindman et al. (2019) demon­ Calvillo, M.P., McLaren, J.W., Hodge, D.O., Bourne, W.M., 2004. Corneal reinnervation
strated large acellular areas on PRK-MMC corneas, replacing the prior after LASIK: prospective 3-year longitudinal study. Invest. Ophthalmol. Vis. Sci. 45,
3991–3996.
α-SMA-positive myofibroblast zone. Although, the nerves continue their Campos, M., Hertzog, L., Garbus, J.J., McDonnell, P.J., 1992. Corneal sensitivity after
path in the acellular zones with an increased nerve density in sub-­ photorefractive keratectomy. Am. J. Ophthalmol. 114, 51–54.
epithelial and epithelial layers compared to corneas that had PRK Cavanagh, H.D., Petroll, W.M., Alizadeh, H., He, Y.G., McCulley, J.P., Jester, J.V., 1993.
Clinical and Diagnostic Use of in Vivo Confocal Microscopy in Patients with Corneal
without MMC. Therefore, MMC enhanced the neurite outgrowth 3
Disease.
months after PRK (Hindman et al., 2019). Cherry, P.M.H., Tutton, M.K., Adhikary, H., Banerjee, D., Garston, B., Hayward, J.M.,
Ramsell, T., Tolia, J., Chipman, M.L., Bell, A., 1994. The treatment of pain following
photorefractive keratectomy. J. Refract. Corneal Surg. 10, S222–S225.
6. Evidence-based treatment of corneal nerves dysfunction
Chuck, R.S., Quiros, P.A., Perez, A.C., McDonnell, P.J., 2000. Corneal sensation after
following refractive surgery laser in situ keratomileusis. J. Cataract Refract. Surg. 26, 337–339.
Darwish, T., Brahma, A., O’Donnell, C., Efron, N., 2007. Subbasal nerve fiber
The treatment of corneal neuropathic disorders related to refractive regeneration after LASIK and LASEK assessed by noncontact esthesiometry and in
vivo confocal microscopy: prospective study. J. Cataract Refract. Surg. 33,
surgery or corneal diseases aims for the relief of patients’ discomfort, 1515–1521.
improvement of the ocular surface, enhanced epithelial healing, and De Oliveira, R.C., Wilson, S.E., 2019. Practical guidance for the use of cyclosporine
nerve regeneration. ophthalmic solutions in the management of dry eye disease. Clin. Ophthalmol. 13,
1115–1122.
Non-preserved artificial tears, lipid and tear film supplementation, Demirok, A., Ozgurhan, E.B., Agca, A., Kara, N., Bozkurt, E., Cankaya, K.I., Yilmaz, O.F.,
proper management of meibomian gland dysfunction (MGD), and 2013. Corneal sensation after corneal refractive surgery with small incision lenticule
punctal plugs can be used to achieve a better ocular surface (Ambrosio extraction. Optom. Vis. Sci. 90, 1040–1047.
Denoyer, A., Landman, E., Trinh, L., Faure, J.F., Auclin, F., Baudoulin, C., 2015. Dry eye
et al., 2008; Galor et al., 2018). Steroids, cyclosporine, and lifitegrast are disease after refractive surgery: comparative outcomes of small incision lenticule
options for control of inflammation (De Oliveira and Wilson, 2019; extraction versus LASIK. Ophthalmology 122, 669–676.
Holand et al., 2017). In addition, the inflammatory decrease can be Diez-Feijóo, E., Grau, A.E., Abusleme, E.I., Duran, J.A., 2014. Clinical presentations of
and causes of recurrent corneal erosion syndrome: review of 100 patients. Cornea
beneficial to treat DED and neuropathic pain. On the other hand, cor­ 33, 571–575.
ticosteroids can delay the neuroregeneration process (Hindman et al., Donnenfeld, E.D., Ehrenhaus, M., Solomon, R., Mazurek, J., Rozell, J.C., Perry, H.D.,
2019). Thus, care should be taken to individualize the use of long-term 2004. Effect of hinge width on corneal sensation and dry eye after laser in situ
keratomileusis. J. Cataract Refract. Surg. 30, 790–797.
corticosteroid treatment due to a possible delay in epithelium healing
Erie, J.C., McLaren, J.W., Hodge, D.O., Bourne, W.M., 2005. Recovery of corneal
(Shahenn et al., 2014). subbasal nerve density after PRK and LASIK. Am. J. Ophthalmol. 140, 1059–1064.
Epithelial healing can be facilitated with bandage contact lenses, Feenstra, R.P.G., Tseng, S.C.G., 1992. Comparison of fluorescein and rose bengal
ointments, and lubricants. Autologous serum and the topical use of staining. Ophthalmology 99, 605–617. https://doi.org/10.1016/s0161-6420(92)
31947-5.
growth factors (NGF, IGF-1, and others) also have a proneural effect Feng, Y.F., Yu, J.G., Wang, D.D., Li, J.H., Huang, J.H., Shi, J.L., Wang, Q.M., Zhao, Y.E.,
(Toda et al., 2018; Levitt et al., 2015). 2013. The effect of hinge location on corneal sensation and dry eye after LASIK: a

9
C.S. Medeiros and M.R. Santhiago Experimental Eye Research 200 (2020) 108243

systematic review and meta-analysis. Graefes Arch. Clin. Exp. Ophthalmol. 251, Marfurt, C.F., Ellis, L.C., 1993. Immunohistochemical localization of tyrosine
357–366. hydroxylase in corneal nerves. J. Comp. Neurol. 336, 517–531.
Frueh, B.E., Cadez, R., Bohnke, M., 1998. In vivo confocal microscopy after Marfurt, C.F., Kingsley, R.E., Echtenkamp, S.E., 1989. Sensory and sympathetic
photorefractive keratectomy in humans. A prospective, long-term study. Arch. innervation of the mammalian cornea. A retrograde tracing study. Invest.
Ophthalmol. 116, 1425–1431. Ophthalmol. Vis. Sci. 30, 461–472.
Gallar, J., Acosta, M.C., Gutierrez, A.R., Belmonte, C., 2007. Impulse activity in corneal Marfurt, C.F., Jones, M.A., Thrasher, K., 1998. Parasympathetic innervation of the rat
sensory nerve fibers after photorefractive keratectomy. Invest. Ophthalmol. Vis. Sci. cornea. Exp. Eye Res. 66, 437–448.
48, 4033–4037. Marfurt, C.F., Cox, J., Deek, S., Dvorscak, L., 2010. Anatomy of the human corneal
Galor, A., 2019. Painful dry eye symptoms: a nerve problem or a tear problem? innervation. Exp. Eye Res. 90, 478–492.
Ophthalmology 126, 648–651. Medeiros, C.S., Marino, G.K., Lassance, L., Thangavadivel, S., Santhiago, M.R., Wilson, S.
Galor, A., Small, L., Feuer, W., Levitt, R.C., Sarantopoulos, K.D., Yosipovitch, G., 2017. E., 2018. The impact of Photorefractive keratectomy and Mitomycin C on corneal
The relationshio between ocular itch, ocular pain, and dry eye symptoms. Trans. Am. nerves and their regeneration. J. Refract. Surg. 34, 790–798.
Ophthalmol. Soc. 115. T5[1-13]. Mietz, H., Addicks, K., Diestelhorst, M., Krieglstein, G.K., 1995. Intraocular toxicity to
Galor, A., Moein, H., Lee, C., Rodriguez, A., Felix, E.R., Sarantopoulos, K.D., Levitt, R.C., ciliary nerves after extraocular application of mitomycin C in rabbits. Int.
2018. Review. Neuropathic pain and dry eye. Ocul. Surf. 16, 31–44. Ophthalmol. 19, 89–93.
Gambato, C., Miotto, S., Cortese, M., Ghirlando, A., Lazzarini, D., Midena, E., 2011. Mietz, H., Prager, T.C., Schweitzer, C., Patrinely, J., Valenzuela, J.R., Font, R.L., 1997.
Mitomycin C-assisted photorefractive keratectomy in high myopia: a long-term Effect of mitomycin C on the optic nerve in rabbits. Br. J. Ophthalmol. 81, 584–589.
safety study. Cornea 30, 641–645. Miller, D.D., Hasan, S.A., Simmons, N.L., Stewart, M.W., 2019. Recurrent corneal
Ganesh, S., Gupta, R., 2014. Comparison of visual and refractive outcomes following erosion: a comprehensive review. Clin. Ophthalmol. 13, 325–335.
femtosecond laser- assisted lasik with smile in patients with myopia or myopic Mohamed-Noriega, K., Riau, A.K., Lwin, N.C., Chaurasia, S.S., Tan, D.T., Mehta, J.S.,
astigmatism. J. Refract. Surg. 30, 590–596. 2014. Early corneal nerve damage and recovery following small incision lenticule
Ganesh, S., Brar, S., Arra, R.R., 2018. Refractive lenticule extraction small incision extraction (SMILE) and laser in situ keratomileusis (LASIK). Invest. Ophthalmol. Vis.
lenticule extraction: a new refractive surgery paradigm. Indian J. Ophthalmol. 66, Sci. 55, 1823–1834.
10–19. Moilanen, J.A., Vesaluoma, M.H., Müller, L.J., Tervo, T.M., 2003. Long-term corneal
Garcia, R., de Andrade, D.C., Teixeira, M.J., Nozaki, S.S., Bechara, S.J., 2016. morphology after PRK by in vivo confocal microscopy. Invest. Ophthalmol. Vis. Sci.
Mechanisms of corneal pain and implications for postoperative pain after laser 44, 1064–1069.
correction of refractive errors. Clin. J. Pain 32, 450–458. Moilanen, J.A., Holopainen, J.M., Vesaluoma, M.H., Tervo, T.M., 2008. Corneal recovery
Hamrah, P., Alipour, F., Jiang, S., et al., 2011. Optimizung evaluation of Lissamine Green after LASIK for high myopia: a 2-year prospective confocal microscopic study. Br. J.
parameters for ocular surface staining. Eye 11, 1429–1434. https://doi.org/ Ophthalmol. 92, 1397–1402.
10.1038/eye.2011.184. Muller, L.J., Pels, L., Vrensen, G.F., 1996. Ultrastructural organization of human corneal
He, J., Bazan, N.G., Bazan, H.E., 2010. Mapping the entire human corneal nerve nerves. Invest. Ophthalmol. Vis. Sci. 37, 476–488.
architecture. Exp. Eye Res. 91, 513–523. Muller, L.J., Vrensen, G.F., Pels, L., Cardozo, B.N., Willekens, B., 1997. Architecture of
Heinz, P., Bodanowitz, S., Wiegand, W., Kroll, P., 1997. In vivo observation of corneal human corneal nerves. Invest. Ophthalmol. Vis. Sci. 38, 985–994.
nerve regeneration after photorefractive keratectomy with a confocal Muller, L.J., Marfurt, C.F., Kruse, F., Tervo, T.M., 2003. Corneal nerves: structure,
videomicroscope. Ger. J. Ophthalmol. 5, 373–377. contents and function. Exp. Eye Res. 76, 521–542.
Hindman, H.B., DeMagistris, M., Callan, C., McDaniel, T., Bubel, T., Huxlin, K.R., 2019. Namavari, A., Chaudhary, S., Yco, L., Jin-Hong, C., Sonawane, S., Khanolkar, V.,
Impact of topical anti-fibrotics on corneal nerve regeneration in vivo. Exp. Eye Res. Sarkar, J., Jain, S., 2012. Neurotrophins and nerve regeneration-associated genes are
181, 49–60. expressed in the cornea after lamellar flap surgery. Cornea 31, 1460e7.
Holland, E.J., Luchs, J., Karpecki, P.M., Nichols, K.K., Jackson, M.A., Sall, K., Tauber, J., Nejima, R., Miyata, K., Tanabe, T., Okamoto, F., Hiraoka, T., Kiuchi, T., Oshika, T., 2005.
Roy, M., Raychaudhuri, A., Shojaei, A., 2017. Lifitegrast for the treatment of dry eye Corneal barrier function, tear film stability, and corneal sensation after
disease: results of a phase iii, randomized, double-masked, placebo-controlled trial photorefractive keratectomy and laser in situ keratomileusis. Am. J. Ophthalmol.
(OPUS-3). Ophthalmology 124, 53–60. 139, 64–71.
Hovanesian, J.A., Shah, S.S., Maloney, R.K., 2001. Symptoms of dry eye and recurrent Patel, S.V., McLaren, J.W., Kittleson, K.M., Bourne, W.M., 2010. Subbasal nerve density
erosion syndrome after refractive surgery. J. Cataract Refract. Surg. 27, 577–584. and corneal sensitivity after laser in situ keratomileusis: femtosecond laser vs
Ishii, R., Shimizu, K., Igarashi, A., Kobashi, H., Kamiya, K., 2015. Influence of mechanical microkeratome. Arch. Ophthalmol. 128, 1413–1419.
femtosecond lenticule extraction and small incision lenticule extraction on corneal Recchioni, A., Sisó-Fuertes, I., Hartwig, A., Hamid, A., Shortt, A.J., Robert Morris, R.,
nerve density and ocular surface: a 1-year prospective, confocal, microscopic study. Vaswani, S., Dermott, J., Cerviño, A., Wolffsohn, J.S., O’Donnell, C., 2020. Short-
J. Refract. Surg. 31, 10–15. term impact of FS-LASIK and SMILE on dry eye metrics and corneal nerve
Jeon, K.I., Hindman, H.B., Bubel, T., McDaniel, T., DeMagistris, M., Callan, C., Huxlin, K. morphology. Cornea 1–7.
R., 2018. Corneal myofibroblasts inhibit regenerating nerves during wound healing. Reinstein, D.Z., Archer, T.J., Gobbe, M., Bartoli, E., 2015. Corneal sensitivity after small-
Sci. Rep. 8, 12945. incision lenticule extraction and laser in situ keratomileusis. J. Cataract Refract. Surg.
Jester, J.V., Brown, D., Pappa, A., Vasiliou, V., 2012. Myofibroblast differentiation 41, 1580–1587.
modulates keratocyte crystallin protein expression, concentration, and cellular light Rodriguez-Prats, J.L., Hamdi, I.M., Rodriguez, A.E., Galal, A., Alio, J.L., 2007. Effect of
scattering. Invest. Ophthalmol. Vis. Sci. 53, 770–778. suction ring application during LASIK on goblet cell density. J. Refract. Surg. 3,
Kauffmann, T., Bodanowitz, S., Hesse, L., Kroll, P., 1997. Corneal reinnervation after 559–562.
photorefractive keratectomy and laser in situ keratomileusis: an in vivo study with a Rosenthal, P., Baran, I., Jacobs, D.S., 2009. Corneal pain without stain: is it real? Ocul.
confocal videomicroscope. Ger. J. Ophthalmol. 5, 508–512. Surf. 7, 28–40.
Kim, W.S., Kim, J.S., 1999. Change in corneal sensitivity following laser in situ Rózsa, A.J., Guss, R.B., Beuerman, R.W., 1983. Neural remodeling following
keratomileusis. J. Cataract Refract. Surg. 25, 368–373. experimental surgery of the rabbit cornea. Invest. Ophthalmol. Vis. Sci. 24,
Kornek, G.V., Haider, K., Kwasny, W., Lang, F., Krauss, F., Hejna, M., Raderer, M., 1033–1051.
Weinlander, G., Depisch, D., Scheithauer, W., 1998. Effective treatment of advanced Sambursky, R., O’Brien, T.P., 2011. MMP-9 and the perioperative management of LASIK
breast cancer with vinorelbine, 5-fluorouracil and l-leucovorin plus human surgery. Curr. Opin. Ophthalmol. 22, 294–303.
granulocyte colony-stimulating factor. Br. J. Canc. 78, 673–678. Santhiago, M.R., Netto, M.V., Wilson, S.E., 2012. Mitomycin C: biological effects and use
Lee, B.H., McLaren, J.W., Erie, J.C., Hodge, D.O., Bourne, W.M., 2002. Reinnervation in in refractive surgery. Cornea 31, 311–321.
the cornea after LASIK. Invest. Ophthalmol. Vis. Sci. 43, 3660–3664. Schimmelpfennig, B., Beuerman, R., 1982. A technique for controlled sensory
Lee, S.J., Kim, J.K., Seo, K.Y., Kim, E.K., Lee, H.K., 2006. Comparison of corneal nerve denervation of the rabbit cornea. Graefes Arch. Clin. Exp. Ophthalmol. 218,
regeneration and sensitivity between LASIK and laser epithelial keratomileusis 287–293.
(LASEK). Am. J. Ophthalmol. 141, 1009–1015. Seiler, T., 1993. Photorefractive keratetcomy: European experience. In: Thompson, F.B.,
Levinson, B.A., Rapuano, C.J., Cohen, E.J., Hammersmith, K.M., D Ayres, B., Laibson, P. McDonnell, P.J. (Eds.), Color Atlas/Text of Excimer Laser Surgery. The Cornea, vols.
R., 2008. Referrals to the wills eye institute cornea service after laser in situ 53–62. Igaku-Shoin, New York, NY.
keratomileusis: reasons for patient dissatisfaction. J. Cataract Refract. Surg. 34, Shaheen, B.S., Bakir, M., Jain, S., 2014. Corneal nerves in health and disease. Surv.
32–39. Ophthalmol. 59, 263–285.
Levitt, A.E., Galor, A., Weiss, J.S., Felix, E.R., Martins, E.R., Patin, D.J., Sarantopoulos, K. Shehadeh-Mashor, R., Mimouni, M., Shapira, Y., Sela, T., Munzer, G., Kaiserman, I.,
D., Levitt, R.C., 2015. Chronic dry eye symptoms after LASIK: parallels and lessons to 2019. Risk factors for dry eye after refractive surgery. Cornea 38, 1495–1499.
be learned from other persistent post-operative pain disorders. Mol. Pain 11, 1–12. Sui, T., Zhang, J., Du, S., Su, C., Que, J., Cao, X., 2014. Potential risk of mitomycin Cat
Li, M., Zhao, J., Shen, Y., Li, T., He, L., Xu, H., Yu, Y., Zhou, X., 2013. Comparison of dry high concentrations on peripheral nerve structure. Neural Regen Res. 9, 821–827.
eye and corneal sensitivity between small incision lenticule extraction and Tervo, K., Latvala, T.M., Tervo, T.M., 1994. Recovery of corneal innervation following
femtosecond LASIK for myopia. PloS One 8, e77797. Photorefractive keratoablation. Arch. Ophthalmol. 112, 1466–1470.
Lin, S.R., Aldave, A.J., Chodosh, J., 2019. Review: recurrent corneal erosion syndrome. Toda, I., 2018. Dry eye after LASIK. Invest. Ophthalmol. Vis. Sci. 59, DES109–DES115.
Br. J. Ophthalmol. 103, 1204–1208. Trabucchi, G., Brancato, R., Verdi, M., Carones, F., Sala, C., 1994. Corneal nerve damage
Linna, T.U., Perez-Santonja, J.J., Tervo, K.M., Sakla, H.F., Alio y Sanz, J.L., Tervo, T.M., and regeneration after excimer laser photokeratectomy in rabbit eyes. Invest.
1998. Recovery of corneal nerve morphology following Laser in situ keratomileusis. Ophthalmol. Vis. Sci. January 35, 229–235.
Exp. Eye Res. 66, 755–763. Tuisku, I.S., Lindbohm, N., Wilson, S.E., Tervo, T.M., 2007. Dry eye and corneal
Linna, T.U., Vesaluoma, M.H., Perez-Santonja, J.J., Petroll, W.M., Alió, J.L., Tervo, T.M. sensitivity after high myopic LASIK. J. Refract. Surg. 23, 338–342.
T., 2000. Effect of myopic LASIK on corneal sensitivity and morphology of subbasal
nerves. Invest. Ophthalmol. Vis. Sci. 41, 393–397.

10
C.S. Medeiros and M.R. Santhiago Experimental Eye Research 200 (2020) 108243

Vehof, J., Sillevis Smitt-Kamminga, N., Nibourg, S.A., Hammond, C.J., 2017. Predictors Wilson, S.E., Medeiros, C.S., Santhiago, M.R., 2018. Pathophysiology of corneal scarring
of discordance between symptoms and signs in dry eye disease. Ophthalmology 124, in persistent epithelial defects after PRK and other corneal injuries. J. Refract. Surg.
280e286. 34, 59–64.
Vestergaard, A., Grønbech, K., Grauslund, J., Ivarsen, A.R., Hjortdal, J.O., 2013. Wolffsohn, J.S., Arita, R., Chalmers, R., Djalilian, A., Dogru, M., Dumbleton, K., Gupta, P.
Subbasal nerve morphology, corneal sensation, and tear film evaluation after K., Karpecki, P., Lazreg, S., Pult, H., Sullivan, B.D., Tomlinson, A., Tong, L.,
refractive femtosecond laser lenticule extraction. Graefes Arch. Clin. Exp. Villani, E., Yoon, K.C., Jones, L., Craig, J.P., 2017. TFOS DEWS II diagnostic
Ophthalmol. 251, 2591–2600. methodology report. Ocul. Surf. 15, 539–557.
Wei, S., Wang, Y., 2013. Comparison of corneal sensitivity between FS-LASIK and Woreta, F.A., Gupta, A., Hochstetler, B., Bower, K.S., 2013. Management of
femtosecond lenticule extraction (ReLEx flex) or small-incision lenticule extraction postphotorefractive keratectomy pain. Surv. Ophthalmol. 58, 529–535.
(ReLEx smile) for myopic eyes. Graefes Arch. Clin. Exp. Ophthalmol. 251, Xia, L., Zhang, J., Wu, J., Yu, K., 2016. Comparison of corneal biological healing after
1645–1654. femtosecond LASIK and small incision lenticule extraction procedure. Curr. Eye Res.
Wilson, S.E., Ambrosio, R., 2001. Laser in situ keratomileusis-induced neurotrophic 41, 1202–1208.
epitheliopathy. Am. J. Ophthalmol. 132, 405–406. Yam, G.H., Williams, G.P., Setiawan, M., Yusoff, N.Z., Lee, X.W., Htoon, H.M., Zhou, L.,
Wilson, S.E., Marino, G.K., Medeiros, C.S., Santhiago, M.R., 2017. Phototherapeutic Fuest, M., Mehta, J.S., 2017. Nerve regeneration by human corneal stromal
keratectomy: science and art. J. Refract. Surg. 33, 203–210. keratocytes and stromal fibroblasts. Sci. Rep. 7, 45396.
You, L., Kruse, F.E., Volcker, H.E., 2000. Neurotrophic factors in the human cornea.
Invest. Ophthalmol. Vis. Sci. 41, 692–702.

11