Secondary Logo

Journal Logo

Laboratory science

Collagen crosslinking of human and porcine sclera

Wollensak, Gregor MD*; Spoerl, Eberhard PhD

Author Information
Journal of Cataract & Refractive Surgery: March 2004 - Volume 30 - Issue 3 - p 689-695
doi: 10.1016/j.jcrs.2003.11.032
  • Free

Myopia is a common ocular disorder that was described by the ancient Greeks. It has been corrected by concave lenses since the 16th century. The prevalence of myopia is around 30% of the general population in the United States and Europe and as high as 60% in Asian countries.1 The prevalence of pathologic myopia with a refractive error greater than −8.0 diopters (D) is around 0.3% of the general population in the United States and 1% in Japan.2

Myopic progression (around −0.5 D in a 2-year interval) in up to 50% of myopes is an unsolved problem.3 Therapeutic attempts to arrest myopic progression comprise nonsurgical treatments such as administration of atropine, tropicamide, pirenzipine, timolol, apomorphine4 and the use of bifocal and multifocal glasses or rigid contact lenses to assist in near work and accommodation. Surgical means of scleral strengthening such as injection of a polymeric composition forming a gel under Tenon's capsule and inducing scar tissue5 or so-called scleral reinforcement operations in which donor sclera, fascia lata, or synthetic bands6,7 are placed around the back of the globe and sutured to the sclera to provide scleral support and prevent axial elongation have shown limited success.

The etiology of myopia is controversial, and a multifactorial cause is assumed. Besides classic risk factors such as genetic predisposition, high level of education, near work, scleral weakness with deranged metabolism of the sclera,8,9 and abnormal scleral collagen10 have been proposed. High myopia, blue sclera, and keratoconus have been described in Ehlers-Danlos syndrome in which a lack of intermolecular crosslinks in collagen is known.11 Conversely, in diabetic patients, where glucose-induced collagen crosslinking is known, axial myopia is rare.12 Moreover, recent animal experiments in tree shrews show the influence of collagen crosslinking on the development of experimental myopia.13

Collagen crosslinking, which was recently introduced for the treatment of progressive keratoconus,14,15 may be a way to strengthen the sclera and prevent myopic progression. In contrast to other treatments, this would correct a cause rather than an effect. In the present in vitro study, we evaluated the efficiency of various methods of collagen crosslinking in human and porcine sclera to increase the biomechanical strength of the sclera. Future animal studies, however, must exclude potential toxic side effects before collagen crosslinking can be applied clinically.

Materials and Methods

Specimen Preparation

Five human eye-bank eyes from nonmyopic subjects with a donor age of 45 to 63 years were obtained within 8 hours post-mortem and were used after the corneal button was removed for transplantation. In addition, 50 porcine cadaver eyes were retrieved within 5 hours post-mortem from the local abattoir. After the episcleral tissue was removed, the anterior segment of the globe was cut off. The opened globe was mounted inversely on a globular-shaped plastic block.

In the porcine eyes, 2 parallel 8.0 mm × 4.0 mm scleral strips were dissected sagitally at the 12 o'clock position (Figure 1) starting 10.0 mm from the limbus using a self-constructed triple-blade scalpel. In the human eyes, 2 parallel 8.0 mm × 4.0 mm scleral strips were dissected in each quadrant beginning 10.0 mm from the limbus, yielding 8 strips per eye. The tissues adherent to the scleral strips internally were carefully peeled off.

Figure 1.
Figure 1.:
(Wollensak) Location of the longitudinally excised porcine scleral strips.


The scleral strips were excised from the equatorial region 10 to 18 mm from the limbus because the scleral thickness there is relatively uniform in human and porcine sclera.15–17

The thickness of the strips was determined using ultrasound pachymetry (Pachette, Technomed).

Crosslinking Treatment

In the human scleras, 4 scleral strips were used per treatment group and compared to the adjacent parallel control strips from the same eye. In the porcine scleras, 10 strips were used per treatment group and compared to the respective parallel control strips from the same eye. All scleral strips including the controls were soaked in 20% dextran solution to avoid hydration-induced swelling. The crosslinking times were chosen according to our previous experience in crosslinking the cornea.15

The following crosslinking treatments were tested:

Physical Crosslinking. (1) Riboflavin−UVA: Ten minutes before the treatment, 0.1% riboflavin photosensitizer solution (10 mg riboflavin-5-phosphate in 10 mL 20% dextran-T-500) was dropped on the treated strips. Ultraviolet A irradiation (370 nm) was applied using 2 double UVA diodes (Roithner Lasertechnik) with an irradiance of 3 mW/cm2 at a distance of 1 cm from the sclera for 30 minutes. (2) Rose bengal and white light: Following application of rose bengal for 10 minutes, the strips were irradiated with a white-light fiber optic for 1 hour.

Chemical Crosslinking. (1) Incubation in modified Eagle medium (MEM) plus glyceraldehyde (0.2 mol) for 5 days. (2) Incubation in glutaraldehyde (0.1%) for 1 hour. (3) Incubation in MEM plus ribose (0.5 mol) for 1 week. (4) Incubation in MEM plus glucose (0.5 mol) for 1 week.

Stress-Strain Measurements

The human and porcine scleral strips were clamped horizontally with 6.0 mm between the jaws of a commercially available microcomputer-controlled biomaterial tester (Minimat, Rheometric Scientific GmbH) (Figure 2). To include the physiological stress range, a prestress of 10 × 103 Pa was used, which required a force of 20 mN in both human and porcine sclera. The strain was increased linearly with a velocity of 1.5 mm min−1, and the stress was measured up to 3.5 × 106 Pa. The stress-strain values were fitted by an exponential function σ = A exp (B × ε) using the SPSS calculation program (SPSS GmbH Software). Young's modulus (E) was calculated for 8% strain as the gradient of the stress-strain graph (E = dσ/dε = A × B exp (B × ε).

Figure 2.
Figure 2.:
(Wollensak) Material tester with scleral strips beween the clamps.

Statistical Evaluation

The stress data necessary for a strain of 8% in the untreated control group and the various treatment groups were compared using an analysis of variance.


Stress-Strain Measurements

The stress at 8% strain was 236.8 ± 179.5 × 103 Pa in untreated porcine scleras and 1037.3 ± 477.6 × 103 Pa in untreated human scleras, demonstrating a 4-fold higher stiffness in natural, untreated human sclera (Tables 1 and 2; Figures 3 and 4). The difference was statistically significant (P = .02, Student t test).

Table 1
Table 1:
Stress and Young's modulus at 8% strain in porcine sclera.
Table 2
Table 2:
Stress and Young's modulus at 8% strain in human sclera.
Figure 3.
Figure 3.:
(Wollensak) Stress-strain curves in porcine scleral strips.
Figure 4.
Figure 4.:
(Wollensak) Stress-strain curves in human scleral strips.

In both porcine and human sclera, a significant increase in corneal rigidity was found after crosslinking by glutaraldehyde, glyceraldehyde, and riboflavin–UVA in decreasing order. The achieved absolute stress values for crosslinked porcine and human sclera were similar, but the relative increase in scleral rigidity was markedly higher in porcine sclera.


The scleral thickness of the human sclera at the equator was 0.4 ± 0.15 mm and of the porcine sclera at the equator, 0.8 ± 0.17 mm.


Efficacious collagen crosslinking with a significant rise in biomechanical strength up to 120% in human sclera and 817% in porcine sclera was achieved after crosslinking by glutaraldehyde, glyceraldehyde, or combined riboflavin−UVA treatment.

Collagen crosslinking is a widespread phenomenon, as seen in aging and diabetes mellitus, and causes mechanical strengthening of collagenous tissue by the induction of intrafibrillar and interfibrillar chemical bonds.18 However, there are differences in the strength of crosslinking among the various crosslinking methods, as shown in the present study. Glutaraldehyde is known to be 1 of the strongest crosslinking agents with 1 of the highest number of crosslinks, and therefore its strong efficiency is not surprising.19 Glyceraldehyde leads to glycation-induced crosslinking, as shown in lens proteins,20 and is used successfully in the crosslinking of gelatin microspheres as pharmaceutical drug-delivery systems.21 Riboflavin−UVA induces physical crosslinking by UVA and the photosensitizer riboflavin and has been used successfully as a relatively strong crosslinking method for the treatment of progressive keratoconus.14 Crosslinking by rose bengal 22 was probably not effective because of poor scleral penetration, which has also been described in corneal stroma.23 Glucose and ribose lead to glycation-induced crosslinking but are relatively ineffective crosslinkers.15,24,25

Effective crosslinkers such as glutaraldehyde and glyceraldehyde led to a relatively higher increase in mechanical stiffness in porcine sclera because of the lower baseline values of untreated porcine sclera, but the maximum stress values after glutaraldehyde and glyceraldehyde crosslinking were remarkably similar in porcine and human sclera. The reason for this phenomenon seems to be a sort of saturation process for these 2 types of crosslinking that leads to almost identical maximum mechanical strength in human and porcine sclera after strong crosslinking, despite different baseline values. However, the lower mechanical strength of untreated porcine sclera (factor 3) is at least partially compensated by greater thickness (factor 2).

The scleral strength and the degree of collagen crosslinking in the sclera seem to play a pivotal role in the development of myopia10 because in myopia one finds axial elongation with a weakened, often staphylomatous, sclera of reduced thickness with a reduction in fibril diameter.26,27 High myopia of up to −18.0 D with thin blue sclera and keratoconus is described in Ehlers-Danlos syndrome in which a lack of reducible intermolecular crosslinks is present in collagen and a deficiency of lysyl oxidase, an enzyme involved in the formation of collagen crosslinks, has been found.11 In diabetic patients, however, where glycation-induced collagen crosslinking is increased, a reduced axial length was found12 and axial myopia is rare.28 Similarly, in animal models, deprivation myopia was increased to −23.0 D6 compared to −11.0 D in a control group by giving monocularly deprived tree shrews systemic injections of the lathyrogenic agent beta–aminopropionitrile, which prevents collagen crosslinking by inhibiting lysyl oxidase.13 In addition, in the myopic tree shrews, the scleral creep rate was increased.29

Therefore, scleral collagen might be a good target for the crosslinking treatment in progressive myopia and should be tried clinically and experimentally10 with the methods we have developed in the present study, ie, glutaraldehyde, glyceraldehydes, and riboflavin–UVA, which proved to be reasonably effective. The somewhat moderate increase in mechanical strength after crosslinking in vitro for human sclera (30% to 120%) should be higher in myopes because they have a significantly lower level of mechanical stiffness with a significantly decreased scleral tensile strength and increased scleral extensibility,8–10,30 as in the untreated porcine scleras, so the degree of crosslinking can be markedly increased by the treatment.

As for the best location of the crosslinking treatment, we know from other studies that in myopia, the pathological changes in the sclera such as thinning, reduction in fibril diameter,26 and increased scleral extensibility are highest at the posterior pole.9,27,29,30 Therefore, the posterior sclera should be the main target of the crosslinking treatment. Special caution, however, must be applied because of the closeness of the macula.

Different modes for in vivo application must be developed for the 3 most effective crosslinking treatments. The sclera has a large and relatively easily accessible surface area and is permeable to molecules as large as 150000 d, as demonstrated in rabbit sclera.31 Therefore, glutaraldehyde (MW 100 d), glyceraldehyde (MW 90.08 d), and riboflavin-5-phosphate (MW 456 d) should easily penetrate the sclera. They could be administered, for example, by peribulbar injection or by direct intrascleral injection after surgical access to the bare sclera had been opened. For repeated or long-term treatments, an osmotic pump,31 a soaked scleral buckle, or less traumatically, repeated parabulbar sub-Tenon's injections as in Avetisov's scleral strengthening injections5 or steroid injections in uveitis could be performed.

To develop practical methods of application and to demonstrate the effect of the crosslinking treatment against the development or progression of myopia, animal studies with form deprivation or optical defocus models in the tree shrew or monkey13 must be performed. Riboflavin–UVA irradiation could be applied only after direct surgical access to the sclera, but the extent of the treatment area should be easily controllable because of the visible UVA-light-induced fluorescence. Glyceraldehyde and glutaraldehyde would have the advantage of a more extended application zone with a larger spatial distribution but reduced spatial control.

Before clinical application, all the crosslinking treatments must be tested for possible toxic effects on the retina, lens, and other ocular structures. Glutaraldehyde is known to have a high cytotoxicity and if used must be tried cautiously as with other cytotoxic agents, for example, 5-fluorouracil or mitomycin in glaucoma filter operations, to avoid damage to the choroid and retina. Glyceraldehyde is usually considered to be nontoxic and is a natural product of metabolism.21,32 Still, specific cytotoxicity to ocular structures must be excluded in animal experiments. Riboflavin–UVA treatment should mainly damage the fibrocytes of the superficial layers of the treated sclera because of the high absorption rate of UVA in sclera.

In summary, we have shown that human and porcine scleral collagen can be crosslinked effectively using glyceraldehyde, glutaraldehyde, and riboflavin–UVA irradiation, leading to a statistically significant increase in the biomechanical rigidity of the sclera. Future in vivo animal and clinical studies must determine the best method of application, exclude side effects, and prove the in vivo effectiveness of the crosslinking treatment in stopping the progression of myopia.


1. Saw S-M, Shih-Yen EC, Koh A, Tan D. Interventions to retard myopia progression in children; an evidence-based update. Ophthalmology 2002; 109:415-427
2. Tano Y. Pathologic myopia: Where are we now? LIX Edward Jackson Memorial Lecture. Am J Ophthalmol 2002; 134:645-660
3. Bullimore MA, Jones LA, Moeschberger ML, et al. A retrospective study of myopia progression in adult contact lens wearers. Invest Ophthalmol Vis Sci 2002; 43:2110-2113
4. Iuvone PM, Tigges M, Stone RA, et al. Effects of apomorphine, a dopamine receptor agonist, on ocular refraction and axial elongation in a primate model of myopia. Invest Ophthalmol Vis Sci 1991; 32:1674-1677
5. Avetisov ES, Tarutta EP, Iomdina EN, et al. Nonsurgical and surgical methods of sclera reinforcement in progressive myopia. Acta Ophthalmol Scand 1997; 75:618-623
6. Jacob JT, Lin JJ, Mikal SP. Synthetic scleral reinforcement materials. III. Changes in surface and bulk physical properties. J Biomed Mater Res 1997; 37:525-533
7. Jacob-LaBarre JT, Assouline M, Conway MD, et al. Effects of scleral reinforcement on the elongation of growing cat eyes. Arch Ophthalmol 1993; 111:979-986
8. Avetisov ES, Savitskaya NF, Vinetskaya MI, Iomdina EN. A study of biochemical and biomechanical qualities of normal and myopic eye sclera in humans of different age groups. Metab Pediatr Syst Ophthalmol 1983-84; 7:183-188
9. Awetissow ES. Unterlagen zur Entstehungstheorie der Myopie. 3. Mitteilung. Die Sklera in der Pathogenese der progredienten Myopie. Klin Monatsbl Augenheilkd 1980; 176:777-781
10. Bell GR. Biomechanical considerations of high myopia: Part I—Physiological characteristics. J Am Optom Assoc 1993; 64:332-338
11. Mechanic G. Crosslinking of collagen in a heritable disorder of connective tissue: Ehlers-Danlos syndrome. Biochem Biophys Res Commun 1972; 47:267-272
12. Løgstrup N, Sjølie AK, Kyvik KO, Green A. Long term influence of insulin dependent diabetes mellitus on refraction and its components: a population based twin study. Br J Ophthalmol 1997; 81:343-349
13. McBrien NA, Norton TT. Prevention of collagen crosslinking increases form-deprivation myopia in tree shrew. Exp Eye Res 1994; 59:475-486
14. Wollensak G, Spoerl E, Seiler T. Riboflavin/ultraviolet-A-induced collagen crosslinking for the treatment of keratoconus. Am J Ophthalmol 2003; 135:620-627
15. Spoerl E, Seiler T. Techniques for stiffening the cornea. J Refract Surg 1999; 15:711-713
16. Olsen TW, Sanderson S, Feng X, Hubbard WC. Porcine sclera: thickness and surface area. Invest Ophthalmol Vis Sci 2002; 43:2529-2532
17. Olsen TW, Aaberg SY, Geroski DH, Edelhauser HF. Human sclera: thickness and surface area. Am J Ophthalmol 1998; 125:237-241
18. Bailey AJ. Structure, function and ageing of the collagens of the eye. Eye 1987; 1:175-183
19. Charulatha V, Rajaram A. Influence of different crosslinking treatments on the physical properties of collagen membranes. Biomaterials 2003; 24:759-767
20. Lee K-W, Simpson G, Ortwerth B. A systematic approach to evaluate the modification of lens proteins by glycation-induced crosslinking. Biochim Biophys Acta 1999; 1453:141-151
21. Vandelli MA, Rivasi F, Guerra P, et al. Gelatin microspheres crosslinked with D,L-glyceraldehyde as a potential drug delivery system: preparation, characterisation, in vitro and in vivo studies. Int J Pharm 2001; 215:175-184
22. Pitts JD, Howell AR, Taboada R, etal. New photoactivators for multiphoton excited three-dimensional submicron cross-linking of proteins: bovine serum albumin and type 1 collagen. Photochem Photobiol 2002; 76:135-144
23. Feenstra RPG, Tseng SCG. Comparison of fluorescein and rose bengal staining. Ophthalmology 1992; 99:605-617
24. Girton TS, Oegema TR, Tranquillo RT. Exploiting glycation to stiffen and strengthen tissue equivalents for tissue engineering. J Biomed Mater Res 1999; 46:87-92
25. Ohan MP, Weadock KS, Dunn MG. Synergistic effects of glucose and ultraviolet irradiation on the physical properties of collagen. J Biomed Mater Res 2002; 60:384-391
26. Liu KR, Chen MS, Ko LS. Electron microscopic studies of the scleral collagen fiber in excessively high myopia. J Formosan Med Assoc 1986; 85:1032-1038
27. Funata M, Tokoro T. Scleral change in experimentally myopic monkeys. Graefes Arch Clin Exp Ophthalmol 1990; 228:174-179
28. Pierro L, Brancato R, Robino X, et al. Axial length in patients with diabetes. Retina 1999; 19:401-404
29. Phillips JR, Khalaj M, McBrien NA. Induced myopia associated with increased scleral creep in chick and tree shrew eyes. Invest Ophthalmol Vis Sci 2000; 41:2028-2034
30. Curtin BJ. Physiopathologic aspects of scleral stress-strain. Trans Am Ophthalmol Soc 1969; 67:417-461
31. Ambati J, Canakis CS, Miller JW, et al. Diffusion of high molecular weight compounds through sclera. Invest Ophthalmol Vis Sci 2000; 41:1181-1185
32. Kosmala JD, Henthorn DB, Brannon-Peppas L. Preparation of interpenetrating networks of gelatin and dextran as degradable biomaterials. Biomaterials 2000; 21:2019-2023
© 2004 by Lippincott Williams & Wilkins, Inc.