Hyperguides

Ophthalmology / Refractive surgery

Corneal Wound Healing after LASIK Surgery

Daniel G. Dawson, MD · Hans E. Grossniklaus, MD · Henry F. Edelhauser, PhD
Article

Daniel G. Dawson, MD · Hans E. Grossniklaus, MD · Henry F. Edelhauser, PhD

Introduction

As refractive surgery has grown in popularity over the last decade, the types of surgical options available to correct a patient’s refractive error have also increased in number and become more complex. A recent annual survey that evaluated the variety and volume of refractive surgery performed by American Society of Cataract and Refractive Surgery (ASCRS) practicing ophthalmologists shows this diversity of surgical options clearly.1 The ACRS survey also continues to report that LASIK is still the number one technique being performed in the United States (approximately 1.5 to 2.0 million LASIK procedures are performed annually in the United States). Although initial excitement about LASIK led primarily to clinical outcome studies, implementation of new cutting-edge technological instruments, and implementation of new adjunctive measures, basic science research on the anatomy, biology, and physiology of the wound healing process after LASIK obviously is still needed to fully understand what transpires postoperatively and what biomechanical alterations occur long-term. Unfortunately, basic science research has lagged behind the clinical advances. Animal–primarily rabbit–and human histopathologic studies were recently completed and appear to answer most of the questions that clinicians and patients have wanted to know about the advantages and disadvantages of LASIK surgery.2-5 Since clinical outcomes and several postoperative complications are directly related to the wound healing process, a basic description of this process is described below.

Anatomy: Histological, Ultrastructural, and Immunofluorescent Studies

Animal and human studies have shown that the cornea heals after LASIK through a series of well-defined stages and steps.2-3

  1. Active wound healing.
    1. Initial epithelial and stromal injury due to LASIK results in apoptosis or necrosis of keratocytes and epithelial cells in and slightly around the wound, and extracellular tissue damage.
    2. Cellular proliferation and migration of surviving keratocytes and epithelial cells occur, which repopulate the acellular wound regions.
    3. Cellular differentiation of keratocytes into one of three distinct morphologic cell types occurs: the migratory keratocyte, the activated keratocyte, or the myofibroblast.
  2. Wound remodeling.
  3. Completion of wound healing.

Although human corneal wound healing is similar to that in animal models, it also has notable differences. Adult human corneas heal less aggressively, more slowly, and not as completely as animal corneas.3-4 Human studies also have shown that LASIK corneal stromal wounds remain in the active wound healing stage up to the first 6 months after surgery before entering the wound remodeling stage that occurs from 6 month to 3 to 4 years after surgery. Wound remodeling is important as it is associated with an improvement in corneal transparency and wound strength.3-5 Theoretically, the LASIK wound should be completely healed 4 years after surgery, but sub-clinical remodeling may continue indefinitely.

Overall, the long-term result in human corneas that have had LASIK is that a hypercellular fibrotic reparative stromal scar develops and persists in wound regions where epithelial-stromal interactions occur (ie, the LASIK flap wound margin), presumably because



chemotactic cytokines and growth factors from epithelial cells, keratocytes, and tears act locally only over short distances where the two tissues are in apposition.2-3 In comparison, a hypocellular primitive reparative stromal scar develops and persists in wound regions where keratocyte injury pathways act by themselves (i.e. the central and paracentral LASIK interface wound, Slide 1).3,4 Thus, only reparative stromal scarring develops after creation of a LASIK flap (by mechanical and femtosecond laser means) and/or excimer laser ablation, but not stromal regeneration. The two stromal wound types have functional significance in that the hypercellular fibrotic stromal scar is strong because it is composed of a dense network of collagen fibrils, but also hazy because of myofibroblastic cells populating this scar type (Slide 2).4 This latter fact was confirmed by
SLIDE 3
immunofluorescent studies that identified alpha-smooth muscle actin positive keratocytes and increased collagen type III staining in the extracellular matrix of the hypercellular fibrotic scar (Slide 2, bottom-left inset).3 In comparison, the hypocellular primitive stromal scar is transparent, but weak in tensile strength, thus explaining its potential to serve as a space for fluid, cells, and microbes to collect (Slide 3).4 A special electron microscopic staining technique supports these latter conclusions as the hypocellular primitive stromal scar was found to primarily consist of abnormally large, nonfibril-bound proteoglycans (Slide 3, bottom-right inset).3 Immunofluorescence studies further showed decreased or absent levels of all normal corneal stromal constituents other than collagen type I in the hypocellular primitive stromal scar (Slide 4).3 An additional variable to consider in this wound healing scheme is that precisely realigned wounds (ie, sutured or unsutured LASIK flap margins with minimal gaping and no epithelial cell plugging) were found to heal more efficiently–stronger and less hazy–than poorly aligned wounds (ie, LASIK flap margins with wide wound gaping, epithelial plugging, or incarceration of adjacent corneal tissue).

The type, location, and amount of stromal scar found in histopathologically studied post-LASIK corneas can be summarized by the following: the hypercellular fibrotic stromal scar at the LASIK flap wound margin averaged 50 µm to 75 µm in depth and 8 µm in thickness and the hypocellular primitive stromal scar at the central and paracentral interface wound (i.e. in the remaining deeper stromal zone away from the epithelial surface) averaged 5 µm in thickness.3 Focal or diffuse epithelial ingrowth at the flap wound margin was found in 53% of post-LASIK corneas that were evaluated by step sections through the entire cornea.3 Post-LASIK corneas with epithelial ingrowth were notably different from those without epithelial ingrowth in that they only had a 5 µm to 10 µm in-depth adjacent zone of hypercellular fibrotic scarring next to the epithelial ingrowth site.


Histopathologic studies performed on human post-LASIK corneas also have shown that all cases had various degrees of focal undulations of Bowman’s layer and overlying focal basal epithelial hypertrophy over the surface of the flap.3 Similarly, at the LASIK flap wound margin, all post-LASIK corneas had focal basal epithelial hypertrophy and, occasionally, epithelial hyperplasia.3 A summary diagram of the wound healing process is shown in Slide 5.

Physiology and Biology: Clinical and Laboratory Studies

Transparency

The normal cornea transmits 95% to 99% of incident visible light because of the stroma’s lattice-like arrangement of collagen fibrils and its similar refractive index changes, and because the cells of the cornea are predominantly transparent.6 Recently, a few studies have suggested that the cells that make up part of the cornea may be the most important determinant to maintaining normal corneal transparency.7 For example, clinical slit-lamp examination and in vivo confocal microscopy of the normal human cornea have shown that most of the back-scattered light (ie, reflected and obliquely back-scattered light) comes from the cellular components in the cornea as opposed to the collagen fibrils or other extracellular matrix constituents (endothelial cells and epithelial cells > nerve cells > keratocytes >> extracellular matrix).7 Similarly, subepithelial stromal haze, which commonly develops after various types of corneal stromal injury at the epithelial-stromal junction (eg, photorefractive keratectomy [PRK], LASIK flap margin, traumatic perforating corneal injury), also appears to result from cellular factors. Wound-response keratocytes have been shown to transform into myofibroblasts--cells that initiate wound contracture and are currently thought to be the primary source of haze after refractive surgery.2-4 A recent ex vivo confocal microscopy study tried to quantitatively measure this back-scattering of light after LASIK surgery.8 The ex vivo confocal study found that 30- µm optical sections of the hypocellular primitive stromal scar back-scattered similar amounts of light as the normal corneal stroma (51.9 vs. 51.8 U), whereas the hypercellular fibrotic stromal scar back-scattered 38% more light than normal (78.4 vs. 57.2 U). The cause of the latter finding was thought to be due to back-scattering actin microfilament stress bundles and/or alterations in cytoplasmic protein levels in myofibroblasts. It has been suggested, but not proven, that myofibroblast persistence and possibly development is caused by biomechanical stress and/or epithelial basement membrane disruption.2,3

Curvature

Postoperative refractive regression is defined as a gradual loss of initial refractive correction. It is thought to result from postoperative corneal wound healing, which causes a change in curvature of the central anterior corneal surface. Postoperative refractive regression appears to occur in human corneas predominantly over the first 6 months after surgery and, after LASIK, averages between 4% and 10% (as opposed to rabbit corneas that regress 100%) (Slide 6).2,3 Compared to LASIK, the amount of regression in human PRK corneas is higher, averaging between 10% and 17% with newer, smoother, large-ablation profiles. The cause for refractive regression after LASIK is multifactorial, but is at least partially due to all of the following: focal epithelial hypertrophy over the center of the LASIK flap, development and growth of two types of stromal scars, and, possibly, an acute anterior forward shift in the cornea (not the same as ectasia).

Based on histopathologic studies, the typical LASIK cornea regresses primarily because, on average, a 5-µm-thick hypocellular primitive stromal scar develops centrally and an 8-µm-thick hypercellular fibrotic stromal scar develops at the flap wound margin, resulting in a re-steepening of the central anterior surface of the LASIK flap after the typical myopic LASIK treatment (Slide 7). The hypercellular fibrotic stromal scar presumably has more of an impact on regression than the central hypocellular primitive stromal scar because it contains contractile myofibroblasts that cinch down on the wound margin 1 to 3 months postoperatively, which further re-steepens the central anterior surface of the LASIK flap (mechanistically similar to how conductive keratoplasty steepens the central anterior corneal surface). Occasionally, the central anterior surface of the LASIK flap may re-steepen more or less than the typical case does. The authors have found that most of these atypical regression cases were caused by poorly aligned flap margins, which sometimes result in focal epithelial hyperplasia or hypoplasia over the center of the flap.


Stability

The biomechanical breaking force of a tissue is called the cohesive tensile strength of the tissue. Normal adult human corneal stroma has been found to have a cohesive tensile strength between 22 g/mm and 36 g/mm, typically being strongest in the peripheral stroma and in the

SLIDE 8A and 8B
anterior third of the central stroma, which are regions of the corneal stroma that have the highest density of bridging collagen filaments and/or interweaving collagen lamellae.5 The weakest measurement typically was found in the central, posterior two-thirds of the stroma. The region of the corneal stroma where the LASIK flap is made (ie, 100 µm-180 µm in depth from the anterior corneal surface) was found to average 30 g/mm (Slide 8A). In contrast, the cohesive tensile strength of the post-LASIK cornea was found to be much weaker than normal corneas, being weakest along the lamellar interface scar, regardless of postoperative time (Slide 8B). Moreover, the mean cohesive tensile strength of the hypocellular primitive stromal LASIK scar averaged 2.4% (0.7 g/mm) of normal controls (30.1 g/mm) and showed no change in strength over time after surgery (Slide 9A); the mean peak cohesive tensile strength of the hypercellular fibrotic stromal LASIK scar gradually increased over time after surgery, reaching maximum values by 3.5 years after surgery when the average was 28.1% (8.5 g/mm) of controls (Slide 9B).5


The latter wound strength study concluded that corneal stromal LASIK wounds were weaker than normal unoperated corneal stroma because normal stromal structures were not regenerated during the wound healing response. Nevertheless, the newly created collagen fibrils in the reparative stromal scars were found to intercalate between and around adjacent old cut fibrils, which is similar to what David Maurice proposed in 1987—that new collagen fibrils in the stromal scar do not reconnect end-to-end with old cut fibril ends.8 Histological and ultrastructural correlations performed in the wound strength study also contributed to our understanding of corneal wound healing as it showed that the amount of densely packed, intercalating collagen fibrils and, probably, the number of myofibroblasts contained in the scar were most indicative of strong scars (30% to 50% of normal strength). On the other hand, stromal scars that contained epithelial cells or non-fibril-bound proteoglycans were generally the weakest (8-15% and 1-5% of normal strength, respectively).

Biomechanical elastic properties (e.g. stress, strain, and Young’s modulus) of the post-LASIK cornea have also recently been evaluated in the laboratory using interferometry techniques.10 Early results (currently only performed in some animal models) suggest that cutting the LASIK flap alone causes reduced corneal strain on the flap (and, thereby, increased corneal strain on the residual stromal bed and, overall, a decreased Young’s modulus over the entire cornea where LASIK took place). This resulted in a 21% increase in the displacement of the central post-LASIK cornea after hydrostatic loading compared to normal corneas. Because the cornea is part of the eye wall and helps contain the intraocular pressure (IOP), the interferometry study provides direct evidence of why post-LASIK corneal ectasia occasionally develops postoperatively and also suggests that the risk for biomechanical instability is higher after LASIK than after PRK or other surface ablative procedures. Until recently, no reliable way to determine the in vivo elastic biomechanical properties of a patient’s cornea existed. The Reichert Ocular Response Analyzer is a new promising tool that appears to determine the in vivo elastic biomechanical properties of a cornea by measuring a property called corneal hysteresis, which represents an aggregate of corneal thickness, rigidity, hydration, and other factors.11

The long-term biomechanical alterations found in a post-LASIK cornea help explain many clinical conundrums. First, these alterations offer an explanation why applanation tonometry IOP readings are lower after LASIK surgery than can be explained by reduced corneal thickness alone (reduced central corneal hysteresis and reduced central corneal Young’s modulus). Secondly, the alterations suggest a reason why corneal edema in post-LASIK corneas results in interface fluid syndrome (the hypocellular primitive stromal scar preferentially absorbs water and the weak cohesive tensile strength of this scar easily swells to form interface fluid pockets). Finally, the alterations also explain why LASIK flaps can be lifted indefinitely or why microbes and cells preferentially accumulate in the central and paracentral interface wound (the weak hypocellular primitive stromal scar heals very poorly resulting in a permanent potential space in all post-LASIK corneas).

Sensitivity

The epithelium of the cornea is described as being the most densely innervated surface epithelium of the body with about 2,500 nerve terminals/mm2 (300 to 600 times more densely innervated than skin).7,12 As the surface area of the adult human cornea is approximately 138 mm2, it has been estimated that each cornea contains 315,000 to 630,000 nerve endings. Most of the nerve fibers in the cornea are sensory in origin and are derived from the ophthalmic branch of the trigeminal nerve (CNV1), which supplies the eye mainly through two long ciliary nerves. Innervation of the cornea (corneal epithelium and the anterior third of the corneal stroma) arises from the long ciliary nerves that penetrate the sclera around the scleral canal and begin to branch in the outermost layers of the choroid as they pass beneath the ora serrata. The main nerve trunks are myelinated and provide branching networks that supply the sclera, episclera, and conjunctiva. The remaining branches form a circumcorneal network around the limbus called the annular plexus where 60 to 80 radial nerve trunks enter the peripheral cornea near or at the midpoint of the corneal stroma. Within 1 mm of the limbus, the stromal nerve trunks lose their myelin sheaths and perineurium, but retain their Schwann cell sheaths. The radial nerve trunks then branch extensively in the anterior third of corneal stroma interconnecting among themselves in the subepithelial plexus. Few, if any, of the nerves pass posteriorly to innervate the area of the deeper stroma and none pass through Descemet’s membrane to supply the endothelial cells.

Although only a few nerve fibers in subepithelial plexus terminate in the anterior third of the corneal stroma, parts of the stromal nerves in the subepithelial plexus have gaps in the Schwann cell sheaths that act as receptor elements. When the nerve bundles of the subepithelial plexus turn 90 degrees to penetrate the Bowman’s layer throughout the peripheral and central cornea, they lose their Schwann cell sheaths. They then again turn 90 degrees and continue parallel to the corneal surface forming the sub-basal nerve plexus, which is composed of nerve bundles that contain straight and beaded nerve fibers. Only the beaded fibers branch from the sub-basal nerve plexus to form epithelial nerves, which course between corneal epithelial cells before finally terminating into 10 to 20 unspecialized free nerve endings. The terminals typically extend up to the most superficial epithelial cell layers. Since different types of epithelial nerve fibers can be distinguished on the basis of their ultrastructure, it appears that corneal epithelial nerves are primarily A-delta and C fibers (touch and pain sensations).

In addition to their sensitivity functions, corneal nerves also appear to be important in maintaining the health of the corneal epithelium and ocular surface via trophic influences and/or other factors. Since the earliest experimental studies by Magendie, it has been shown that dysfunction of corneal innervation produces a degenerative condition to the corneal epithelium called neurotrophic keratitis.2,5,13 Although many disorders can cause neurotrophic keratitis, LASIK has recently gained the most attention for temporarily causing this condition.14 This usually results in short-term symptoms of dry eye after refractive surgery, which is not only due to a transient decrease in corneal sensation and epithelial trophic factors, but also from a transient decrease in tear production.2 Dry eye symptoms have been found to be more frequent, more severe, and longer in duration following LASIK than following PRK, because the total surface area and depth of corneal nerve injury are greater after LASIK than PRK. Additionally, re-innervation takes more time with LASIK than with PRK. In most cases, LASIK is immediately followed by loss of corneal sensation over the flap and gradual disappearance of most of the corneal nerves in the flap over the first 2 days after surgery. Significant re-growth of the cut nerves starts between 1 and 3 months after the procedure with corneal sensation typically recovering to normal levels only by 6 to 12 months postoperatively. However, the total length, morphology, and sensation of the re-grown corneal nerve fibers never completely return to normal preoperative levels and only reaches maximum levels by 1 to 2 years after LASIK. The temporary de-nervation of the post-LASIK cornea also has been associated with a transient decrease in anterior stroma cellular density, which may be due to Schwann cell degeneration and/or loss of neurotrophic influences on keratocytes.3

Conclusion

Although the enthusiasm associated with LASIK surgery is high due to its good clinical success in permanently correcting refractive errors, basic science research has now more clearly shown that LASIK, like most surgical procedures, has both short-term and long-term advantages and disadvantages. After considering all these factors, a prospective LASIK patient, with the help of the ophthalmologist, can decide whether LASIK surgery is in his or her best interest.

References

  1. Leaming DV. Practice styles and preferences of ASCRS members – 2003 survey. J Cataract Refract Surg. 2004; 30:892-900.
  2. Netto MV, Mohan RR, Ambrosio R, et al. Wound healing in the cornea. A review of refractive surgery complications and new prospects for therapy. Cornea. 2005; 24:509-522.
  3. Dawson DG, Kramer TR, Grossniklaus HE, Waring GO III, Edelhauser HF. Histologic, ultrastructural, and immunofluorescent evaluation of human laser-assisted in situ keratomileusis corneal wounds. Arch Ophthalmol. 2005; 123:741-756.
  4. Dawson DG, Edelhauser HF, Grossniklaus HE. Long-term histopathologic findings in human corneal wounds after refractive surgical procedures. Am J Ophthalmol. 2005; 139:168-178.
  5. Schmack I, Dawson DG, McCarey BE, et al. Cohesive tensile strength of human laser-assisted in situ keratomileusis wounds with histogic, ultrastructural, and clinical correlations. J Refract Surg. In press.
  6. Maurice DM. The structure and transparency of the cornea. J Physiol. 1957; 136:263-286.
  7. Dawson DG, Watsky MA, Geroski DH, Edelhauser HF. Cornea and sclera. In: Duane’s Foundation of Clinical Ophthalmology. Lippincott Williams & Wilkins, Philadelphia, PA. In press.
  8. Dawson DG, Holley GP, Geroski DH, et al. Ex vivo confocal microscopy of human LASIK corneas with histologic and ultrastructural correlation. Ophthalmol. 2005; 112:634-644.
  9. Maurice DM. The biology of wound healing in the corneal stroma. Castroviejo lecture. Cornea. 1987; 6:162-168.
  10. Jaycock PD, Lobo L, Ibrahim J, Tyrer J, Marshall J. Interferometric technique to measure changes in the cornea induced by refractive surgery. J Cat Refract Surg. 2005; 31:175-184.
  11. Luce DA. Determining in vivo biomechanical properties of the cornea with an ocular response analyzer. J Cat Refract Surg. 2005; 31:156-162.
  12. Muller LJ, Marfurt CF, Kruse F, Tervo TMT. Corneal nerves: structure, contents, and function. Exp Eye Res. 2003;76:521-542.
  13. Magendie J. De La influenca de la cinquime paire nerfts surla nutrition et les functions de loeil. J Physiol. 1824;4:176-182.
  14. Wilson SE, Ambrosio R. Laser in situ keratomileusis-induced neurotrophic epitheliopathy. Am J Ophthalmol. 2001; 132:405-406.