Is scleral cross-linking a feasible treatment for myopia control?


  • *Both authors contributed equally to this paper.

The question of how eyes enlarge to become myopic has not received the attention warranted it, given that it defines the scope of possible treatment options to control myopia progression. Nonetheless, some aspects of eye growth regulation now seem beyond dispute: (1) most of the regulation occurs locally, driven by the retina in response to altered visual experience, and (2) the sclera undergoes increased extracellular matrix remodeling during the accelerated eye growth that underlies myopia, with the net result on scleral ultrastructure being a decrease in the average diameters of collagen fibers and biomechanical weakening. These changes are exaggerated in highly myopic eyes. Scleral buckling is increasingly being employed as a prophylactic strategy directed at preventing the complication of high myopia, which include scleral staphylomas and associated retinal complications, yet is both highly invasive and dependent on access to healthy donor scleral tissue. The rising prevalence of high myopia calls for new treatment options. A logical question is whether the cross-linking technology currently being employed to strengthen the corneal stromal matrix to treat similar problems of biomechanical weakness in the cornea, e.g., keratoconus, can be used to treat myopia, specifically to slow and/or arrest its progression. Is strengthening of the sclera sufficient to stop myopia progression or are there other over-riding influences to be accounted for? And what are the obstacles to adopting this technology to develop myopia control therapies?


The importance of myopia not only stems from its wide prevalence and consequent large cost to health services, but from the association of high degrees of myopia (>6D) with potentially blinding pathologies such as retinal detachment, glaucoma and macular degeneration. In the great majority of cases (>95%), myopia is associated with abnormal postnatal eye growth, leading to an enlarged vitreous chamber depth and hence excessive ocular axial length; changes in the refractive powers of the lens and cornea contribute minimally.[1] The enlarged vitreous chamber is associated with thinning and localized ectasia of the posterior sclera, as well as thinning of collagen fibre bundles and reduction in the size of individual collagen fibrils. In addition to the associated weakening of the posterior sclera, which may cause staphyloma with possible detrimental effects on vision, the increased eye size in myopia has been found to subject the retina and choroid to higher mechanical stresses,[2] increasing the risk of secondary chorio-retinal pathologies.

The combination of enlarged scleral cup in high myopia and chorio-retinal pathology has been addressed by a number of surgical, sclera buckling techniques to control myopia progression, arrest posterior sclera deformation and treat retinal detachment. The techniques, which use either donor sclera tissue[3] or specially designed polymeric devices,[4] have been effective in critical cases of high myopia and achieve reasonable success rates. However, the techniques suffer from considerable drawbacks in that they are highly invasive, require general anaesthetics and some of them rely on the availability of healthy scleral donor tissue, which is becoming increasingly difficult to obtain with the rising prevalence of myopia.

Earlier mechanical studies have shown that scleral tissue of myopic, enlarged eyes have almost the same elasticity as emmetropic eyes. However, they also indicated that myopic scleras have different time-dependent viscoelastic properties that would allow them to creep (deform under a constant pressure) over time. This characteristic means that, given sufficient time, intraocular pressure loading can induce the ocular enlargement seen in myopia.[5] It has also been reported that the creep and change in form observed in high myopia were associated with a reduction in intermolecular collagen crosslinks.[6, 7] This association was further supported by the fact that in diabetes, where glucose-induced collagen crosslinking is known, high myopia is rare.[8]

These observations indicate that techniques that induce collagen crosslinking and change scleral viscoelastic behaviour in a way that reduces its creep over time, would be expected to retard or inhibit eye elongation, slow the progression of myopia and consequently reduce the development risks of associated sight threatening pathologies. One such technique is based on the combined use of riboflavin and UVA radiation, which has been found successful in crosslinking the cornea and producing stable and considerable increases in its mechanical stiffness and strength.[9] While the technique has been successful in restoring clear vision in keratoconus,[10] its potential for application in the sclera is in doubt as it will require surgical exposure of the sclera and may present a cytotoxic risk to the retina. Earlier research reported considerable loss in the photoreceptors and the retinal pigment epithelium with the combined use of riboflavin and UVA in rabbit sclera.[11] The research recommended reducing the irradiation dose below the cytotoxic level of the retina and testing the method in a myopia animal model before considering its clinical implementation.

An alternative, and possibly a more suitable approach, is based on chemical crosslinking by glycation using glyceraldehyde, which requires a less invasive application procedure and has been shown to increase scleral stiffness by close to 150%.[12] In addition, unlike other crosslinking agents such as glutaraldehyde, glyceraldehyde is a natural product of metabolism and therefore carries no cytotoxicity risk.[13] The crosslinking produced by glyceraldehydes is expected to compensate for the lack of reducible intermolecular crosslinks in collagen, and the deficiency of lysyl oxidase, an enzyme involved in the formation of collagen crosslinks–both of which have been found in high myopia.[6] However, while this technique appears to have a significant promise in retarding progression of myopia, it still requires more research to optimise its method of application, assess its biochemical effects on ocular and surrounding tissues, and to understand its long-term mechanical effectiveness.


The high prevalence of myopia worldwide and the sight-threatening consequences of the progressive eye enlargement in high myopia are well documented.[14] There is clearly a pressing need to control myopia, both by reducing the incidence of myopia within the population and also by reducing the progression of myopia in those who develop it. Reducing an individual's myopia progression could decrease the severity of their myopia later in life and thus reduce the risk of vision loss. The morbidity of high myopia results from stretching and damage to the retina and choroid caused by expansion of the scleral coat at the posterior pole. Scleral expansion is associated with thinning of the sclera and in extreme cases formation of staphyloma (local ectasia) at the posterior pole.

Recently, collagen cross-linking (CXL) has emerged as a novel and promising approach to the management of ocular ectasias. For example, clinical studies suggest that CXL of the cornea is particularly effective in stabilizing progressive keratoconus. The most common CXL procedure uses riboflavin and ultraviolet-A (UVA) light exposure to increase the biomechanical stiffness of the corneal tissue by increasing cross-linking between stromal collagen fibres. The increased rigidity of the tissue stabilises the corneal ectasia and leads to improved clinical outcomes in keratoconus.[15] It has been proposed[16] that a similar process (cross-linking the collagen fibres of the sclera) might be employed to increase the stiffness of the sclera and reduce the progressive enlargement of the vitreous chamber and so inhibit myopia progression. Studies of experimental myopia in animals show that scleral tissue from myopic eyes has reduced mechanical stiffness and also increased creep compliance compared to sclera from normal eyes.[17] Scleral CXL in vitro can increase the mechanical stiffness of normal (non-myopic) animal[18] and human[19] scleral tissue although the effect on creep compliance has not been assessed. However, animal studies also suggest a role for scleral myofibroblast cells in determining the mechanical properties of sclera.[20] Data on human myopic sclera is sparse and it remains unclear whether altered mechanical properties of the collagen matrix alone is the central causative factor in the development of human myopia, or whether the cellular components such as myofibroblasts also play an important active role.

There are several reasons to suggest that scleral CXL is unlikely to be a successful treatment for myopia control in the foreseeable future. The reasons relate to the natural course of myopia progression and also to the anatomical location and nature of the relevant scleral tissue. Myopia typically develops and undergoes most rapid progression in children in whom normal growth of the eye, including its refractive components, is still occurring. Thus in children, application of CXL would be expected to inhibit both the normal and abnormal scleral expansion, which may have unpredictable consequences for refractive development.[21] Myopia progresses most rapidly in its early stages, and therefore scleral CXL, like all myopia control treatments would be most effective when applied as early as possible. In the early stages and on a case-by-case basis, predicting which children will progress to high myopia is notoriously unreliable. Several studies show that CXL based on riboflavin/UVA damages the cellular components of tissue,[22] and so scleral CXL is unlikely to be employed until a patient has demonstrated significant myopia progression. Moreover, there are important differences between corneal and scleral tissue which suggest that even if used to treat progressive expansion of a staphyloma, CXL techniques may not be as readily transferable from cornea to sclera as has been suggested. The cornea is thin, transparent and avascular, whereas the sclera is thicker, opaque tissue with its own, albeit sparse vasculature, penetrated by blood vessels supplying the choroid. Blood vessels, including capillaries, are composed of important cellular components as well as basement membranes including collagen. Conventional CXL is likely to damage the cellular components of blood vessels and capillaries within scleral tissue. Moreover, all CXL techniques, whether they are toxic to cells or not,[21] will almost certainly also cross-link the collagen in blood vessels and capillaries, which would permanently alter their material properties with potentially harmful consequences. In addition, if scleral myofibroblasts play a role in determining the mechanical properties of sclera,[20] then they will also be damaged by conventional CXL procedures. The linkages between myofibroblasts and surrounding tissue are also likely to be altered by collagen cross-linking. A further consideration is the anatomical location of the relevant scleral tissue comprising the posterior wall of the eye. Compared to the cornea, this region is difficult to access and difficult to visualise while applying treatment. Moreover this region, particularly around the optic nerve head, is the site at which the long and short posterior ciliary arteries pass through the sclera to supply blood to the ciliary body and outer retina via the choroid. Application of CXL to this region would almost inevitably affect the cells and collagen in the walls of blood vessels supplying the choroid and could potentially spill-over to affect the choroidal vessels themselves.

Although corneal CXL appears relatively safe, complications have been reported and failed corneal CXL is typically resolved with penetrating keratoplasty. In contrast, resolving failed scleral CXL, possibly resulting from cell or vessel damage, would pose serious problems. Alternative posterior scleral reinforcement techniques[23] are in their infancy compared to the highly developed techniques of keratoplasty.

In conclusion, it seems unlikely that scleral CXL would provide a successful treatment for myopia control. CXL may damage the cellular components of tissue, and although CXL aims to alter the stiffness of the sclera by cross-linking collagen, there is evidence that in life the cellular components of the sclera are also important in determining scleral properties. In children, CXL is likely to interfere with the eye-enlargement of normal eye growth, with unknown consequences for refractive development. Apart from possible cell damage, collagen cross-linking is also likely to occur in the blood vessels supplying the choroid as well as in scleral vessels, with the risk of permanent damage to choroidal and scleral tissue. Such considerations need careful investigation before contemplating long term alterations to such a ubiquitous component of tissue as collagen.


Christine Wildsoet

Center for Eye Disease & Development, School of Optometry, University of California, Berkeley, USA


Points of agreement

The high prevalence of myopia world-wide and sight-threatening complications of high myopia leave no doubt about the need for effective interventions. The two contributing authors are in agreement on the role of altered sclera biomechanical (viscoelastic) properties to the morbidity of high myopia, through continued creep (localized or more general) and thus stretching of the adjacent choroid and retina.

Scleral buckling, applied at the posterior pole using donor sclera to stabilize the weakened sclera, cannot be considered new technology, and did not receive strong support from either author. It was considered in its infancy by one author and the other questioned its use for other than critical cases based the invasive nature of the procedure and also the questionable availability of healthy donor scleral tissue long-term.

Neither doubted that collagen crosslinking techniques are able to stiffen and stabilize collagen-based tissues such as the sclera, given the confirmed success of the UVA radiation - riboflavin combination treatment for a variety of corneal conditions including keratoconus. Nonetheless, it is clear to this editor that many important questions remain to be answered before scleral application of such technology reaches the clinic.

Points requiring further study

Key questions that need to be resolved before collagen crosslinking technologies are applied to the sclera to treat myopia, include:

  • What role do myofibroblasts play in scleral physiology and biomechanics, and what are the effects of crosslinking technologies on their interactions with sclera collagen;
  • What is the nature of the cytotoxic reactions to such technologies - what ocular tissues are affected and are the effects reversible; can glyceraldehye glycation crosslinking achieve the same treatment effect without cytotoxicity or can the UVA radiation/riboflavin treatment treatment protocol be manipulated to avoid adverse effects;
  • What are the implications of incidental collagen crosslinking within the walls of blood vessels passing through the sclera and in the nearby choroid;
  • Who is the target population – should children at-risk of developing high myopia be treated and if so, how will they be identified, and what is the risk of inappropriate treatment, resulting in adverse effects of normal ocular growth;
  • What is the back-up plan for failed treatments (for corneal treatments, keratoplasty serves that purpose).


  • Image of creator

    Ahmed Elsheikh

  • Image of creator

    John R. Philips