|Home | About | Journals | Submit | Contact Us | Français|
To characterize scleral biomechanics in both eyes of eight monkeys in which chronic intraocular pressure (IOP) elevation was induced in one eye.
Each posterior sclera was mounted on a pressurization apparatus, IOP was elevated from 5 to 45 mm Hg while the 3D displacements of the scleral surface were measured by speckle interferometry. Finite element (FE) models of each scleral shell were constructed that incorporated stretch-induced stiffening and multidirectionality of the collagen fibers. FE model predictions were then iteratively matched to experimental displacements to extract unique sets of scleral biomechanical properties.
For all eyes, the posterior sclera exhibited inhomogeneous, anisotropic, nonlinear biomechanical behavior. Biomechanical changes caused by chronic IOP elevation were complex and specific to each subject. Specifically: (1) Glaucomatous eyes in which the contralateral normal eyes displayed large modulus or thickness were less prone to biomechanical changes; (2) glaucomatous scleral modulus associated with an IOP of 10 mm Hg decreased (when compared with that of the contralateral normal) after minimal chronic IOP elevation; (3) glaucomatous scleral modulus associated with IOPs of 30 and 45 mm Hg increased (when compared with that of the contralateral normal) after moderate IOP elevation; and (4) FE-based estimates of collagen fiber orientation demonstrated no change in the glaucomatous eyes.
Significant stiffening of the sclera follows exposure to moderate IOP elevations in most eyes. Scleral hypercompliance may precede stiffening or be a unique response to minimal chronic IOP elevation in some eyes. These biomechanical changes are likely to be the result of scleral extracellular matrix remodeling.
Glaucoma is the second leading cause of blindness worldwide1,2 and leads to vision loss through irreversible damage to retinal ganglion cell axons as they pass through the scleral canal at the optic nerve head (ONH). Although glaucoma pathogenesis is not well understood and could also involve direct damage to the retinal ganglion cells or lateral geniculate, the biomechanical environment of the ONH has been hypothesized to play an important role in the neuropathy. The sclera is an important factor in ONH biomechanics, and recent work strongly suggests that the biomechanics of the posterior sclera and lamina cribrosa are tightly coupled.3–6 The sclera is the stiffest tissue of the eye and provides the mechanical boundary conditions for the lamina cribrosa at its insertion into the scleral canal wall. Computational models have shown that scleral stiffness7 and scleral collagen fiber organization3,4 dictate the IOP-induced deformation exhibited by the ONH. Of note, the sclera exhibits significant biomechanical changes with age8 and after exposure to chronic IOP elevation, as shown herein, both of which are major risk factors in glaucoma.
To investigate scleral biomechanics as a factor in glaucomatous damage, it is important to quantify its mechanical response to IOP. The stresses and strains in the sclera are significant at normal IOPs,8,9 and exposure to elevated IOP could lead to changes in scleral biomechanics that precede the onset of glaucoma or occur very early in disease progression. We therefore focused this work on the scleral changes that are present at the early stages of the disease in nonhuman primates, which could indicate that scleral biomechanics and its IOP-induced changes contribute to both the individual susceptibility to glaucoma and its pathogenesis.
We recently described our ex vivo method to experimentally measure the 3-D deformation pattern and thickness of posterior sclera from both eyes of four young and four old monkeys after IOP elevation from 5 to 45 mm Hg.8 In that report, we modeled the posterior sclera as a nonlinear, anisotropic, inhomogeneous soft-tissue using a fiber-reinforced constitutive theory that included stretch-induced stiffening and multidirectionality of the collagen fibers.3 We then derived eye-specific sets of scleral biomechanical properties based on the experimental observations using an inverse finite element (FE) method.9 Our results showed that posterior and peripapillary sclera in old monkeys is significantly stiffer than that from young monkeys and is therefore subject to higher stress but lower strain at all levels of IOP. Whether age-related changes in scleral biomechanics contribute to the development and progression of glaucoma remains to be determined.
We have demonstrated significant alterations in the architecture of the laminar and peripapillary scleral connective tissues at the earliest stage of experimental glaucoma (EG) in the monkey eye.10–18 In the lamina cribrosa of EG monkey eyes, permanent deformation and thickening,16 accompanied by increases in both total connective tissue volume and total number of laminar beams, were observed.15 In addition, we observed posterior deformation accompanied by thinning in the peripapillary sclera in the EG eyes.13,14 These morphologic changes are likely to result from both physical deformation in response to the applied load (IOP) and connective tissue remodeling in response to chronically elevated stress and strain levels.15,19,20 Glaucomatous remodeling of the connective tissues by optic nerve head (ONH) astrocytes and scleral fibroblasts may in turn contribute to retinal ganglion cell axonal damage through direct and/or indirect cellular mechanisms.21,22
We have shown that the biomechanical properties of the peripapillary sclera are altered in early EG using uniaxial testing and linear viscoelastic theory.11 That study demonstrated a significant increase in the equilibrium modulus of the peripapillary sclera from glaucomatous monkey eyes, but no changes were seen in the time-dependent viscoelastic parameters in those eyes.
In the present study, we used novel three-dimensional (3-D) experimental and computational techniques3,8,9 to characterize peripapillary and posterior scleral biomechanics in both eyes of eight adult monkeys in which one eye had been exposed to chronic, laser-induced IOP elevations of modest to substantial magnitude and duration.
All experiments adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. In this study, EG was induced in one eye of eight rhesus monkeys that had no initial ocular abnormality. Our protocol for generating and detecting laser-induced EG has been fully described in previous reports.12–14,16 Briefly, both eyes of each monkey underwent IOP measurement (Tonopen XL; Reichert, Depew, NY) and confocal scanning laser tomographic (CSLT) imaging (Heidelberg Retinal Tomograph; Heidelberg Engineering, Heidelberg, Germany) at manometrically controlled IOP of 10 mm Hg on three to five occasions while normal. Then, imaging was performed every 1 to 2 weeks after laser photocoagulation of the trabecular meshwork (OcuLight TX; Iridex, Mountain View, CA) in one eye, until the onset and/or progression of CSLT-detected ONH surface change by mean position of the disc (MPD) or HRT II topographic change analysis (TCA).23 Four monkeys were observed until the onset of MPD change (confirmed twice), which is our existing criterion for onset of early experimental glaucomatous damage to the ONH.24 Four additional monkeys were observed up to MPD onset and at least one episode of further progression (confirmed twice) to achieve a more advanced stage of ONH (and presumably scleral) alteration. All the monkeys were killed after an intramuscular injection of ketamine/xylazine with an intravenous injection of pentobarbital sodium.
The load-bearing connective tissues of the posterior sclera directly experience glaucomatous alterations of intraocular pressure,11 whereas they are likely to be only minimally affected by glaucomatous alterations to the neural and connective tissues of the ONH. We therefore chose to stage the animals by EG eye cumulative IOP insult rather than by the magnitude of MPD or TCA ONH surface change. We defined cumulative IOP insult for each EG eye to be the difference between the area (computed using the trapezoidal rule) under its IOP-time curve (Fig. 1, red) minus that of its contralateral normal eye (Fig. 1, blue;). Monkey 1 therefore received the lowest cumulative IOP insult and monkey 8, the highest (Table 1).
For this study, we used an experimental protocol for pressure testing the posterior scleral shell that has been fully described in our other reports.8,9 Briefly, all monkeys were anesthetized with an intramuscular injection of ketamine/xylazine and killed with an intravenous injection of sodium pentobarbital. The eyes were enucleated immediately after death, and the posterior scleral shells were cleaned of intra- and extraorbital tissues and individually mounted on a custom-built pressurization apparatus (see Ref. 9 for a detailed description of the apparatus). After mounting, each shell was blotted dry, covered with a white contrast medium (ProCAD Contrast Medium; Ivoclar, Schaan, Lichtenstein), and immediately immersed in isotonic saline at room temperature. IOP was incrementally increased from 5 to 45 mm Hg, and the resulting full-field 3-D displacements of each shell surface were recorded at a resolution of 0.1 μm with an electronic speckle pattern interferometry sensor25,26 (Q100; Ettemeyer AG, Ulm, Germany). Internal pressure was then reset to 5 mm Hg, and scleral topography was manually measured with a 3-D digitizing arm (MicroScribe G2X; Immersion, San Jose, CA). Scleral thickness was also measured at 20 predetermined locations (see Ref. 9 for details) with a 20-MHz ultrasound transducer (PacScan 300P; Sonomed, Inc., Lake Success, NY). The topography and thickness data were combined to create an eye-specific reference geometry for the posterior scleral shell in each subject.
Our constitutive theory for simulating the biomechanical behavior of the posterior sclera and our inverse FE method of extracting mechanical properties based on the scleral pressure testing outlined above have been fully described in our previous reports.3,8,9 Both are briefly explained as follows.
We modeled the posterior sclera as a nonlinear, anisotropic, inhomogeneous soft tissue by using fiber-reinforced composite theory. Within this framework, we assume that scleral collagen fibers provide most of the resistance to IOP and are responsible for two important biomechanical characteristics (nonlinearity, described with the parameters c3 and c4, and anisotropy, with θp and k; Fig. 2). The model assumes that scleral collagen fibers are embedded within a ground substance matrix, not to be confused with the extracellular matrix, which includes all the remaining tissue constituents (e.g., water, fibroblasts, proteoglycans, and elastin). Hence, the ground substance matrix is of low stiffness compared with the collagen fibers, and its mechanical behavior is characterized with a single intrinsic parameter, c1, as isotropy is assumed (i.e., no preferred orientation).
Using our inverse FE method, we estimated a set of thirteen model parameters (c1, c3, c4, k1, k2, and θp1 to θp8; Fig. 3) from the scleral mechanical testing data for each posterior scleral shell. This strategy uses a genetic optimization algorithm (differential evolution47–49) to generate a unique value for each of the 13 model parameters by simultaneously varying all 13 parameters until the FE-predicted displacements closely match the experimental deformations at five IOP levels simultaneously (7, 10, 20, 30, and 45 mm Hg).
Two additional measures of scleral stiffness were derived from the 13 output parameters: the tangent modulus and the structural stiffness. Because of their high clinical relevance, these two stiffness measures are emphasized in the Results and Discussion sections.
The model parameters c3 and c4 describe the stretch-induced stiffening of the collagen fibers, but do not represent a direct measure of scleral stiffness as a function of IOP. To capture scleral stiffness, the tangent modulus along the preferred fiber orientation was calculated and mapped. The tangent modulus is an estimate of local scleral material stiffness, varies with IOP, and is independent of geometry.9
Clinical discussions concerning the observable mechanical behavior of the corneoscleral shell often confuse the difference between the biomechanical material properties of the tissue (the stiffness of the bulk material itself), its architecture (thickness), and the resultant net structural behavior. Although the material properties of the tissues are intrinsic to the material itself, the structural stiffness of the shell incorporates both the material properties and geometry of a complex load-bearing structure into a composite measure of the structure's resistance to deformation. We defined the structural stiffness of the sclera as the local product of the scleral thickness and the tangent modulus to capture the mechanical behavior of the scleral shell as a structure in response to acute IOP elevations.
Regional thickness differences between the normal and glaucomatous eyes of each animal were compared to the regional thickness differences between bilaterally normal eyes of eight previously studied monkeys,8 by using a multiple linear regression.
Posterior displacements between the two eyes of each monkey could not be directly compared, because all displacements were computed with respect to the position of the scleral clamp, which varied to a certain degree among eyes. However, the derived output measures (stress, strain, and structural stiffness) can be directly compared because they are unaffected by scleral clamp position.
Tangent modulus and structural stiffness distributions for the eight adult normal eyes were compared to the previously reported data for four young and four old bilaterally normal monkeys at IOPs of 10, 30, and 45 mm Hg using the general estimating equation method as used in our previous report.8
Regional distributions of maximum principal stress, maximum principal strain, tangent modulus, and structural stiffness were compared between the normal and glaucomatous eyes of each monkey at IOPs of 10, 30, and 45 mm Hg by paired t-tests. Note that the regional data for these output measures (i.e., stress, strain, modulus, and stiffness) were not normally distributed, which violates the assumptions underlying t-tests. To address this, and as commonly performed using the central limit theorem, 3000 means were calculated for each parameter, each consisting of 300 randomly sampled values from the original distribution. The values of 300 and 3000 were chosen as the minimum values necessary for the t-statistic values to be unaffected by non-normality. These normally distributed means were then used in the t-tests to compare the parameters for normal and EG eyes. For each output measure (i.e., stress, strain, modulus, and stiffness), alterations due to chronic IOP elevation were deemed to be significant for that particular monkey when the t statistic from comparing the normal and glaucomatous eye of that monkey exceeded all t statistics computed from comparisons of bilaterally normal eyes of eight other monkeys.8
Finally, anisotropy parameters (i.e., preferred fiber orientation and fiber concentration factor) were compared between groups by paired t-tests in both the peripapillary and peripheral sclera.
Scleral thickness, experimentally measured at an IOP of 5 mm Hg at 20 locations and interpolated over each scleral shell, is shown in the first row of Figure 4 and in Figure 5 (averaged values). In all eyes, the peripapillary sclera was much thicker (276–498 μm on average in the normal eyes) than the peripheral sclera (157–252 μm on average in the normal eyes; P < 0.0001). Among those regions considered (inferior, inferonasal, nasal, superonasal, superior, superotemporal, temporal, and inferotemporal), the sclera was significantly thicker in the temporal region for all 16 eyes (P < 0.001).
For each eye, good agreement was obtained between the FE-predicted and experimentally measured posterior displacements between 5 and 30 mm Hg (second and third rows of Fig. 4; Table 1; maximum averaged nodal error, 4.08 μm). Although not all cases are included in Figure 4, good agreement was also obtained for each IOP range (5–7, 5–10, 5–20, 5–30, and 5–45 mm Hg) and for all displacement components, suggesting that our constitutive model was able to accurately capture the complex 3-D deformations of the monkey peripapillary and posterior sclera after acute elevation of IOP.
For each eye, the sclera stiffened considerably as IOP increased, as demonstrated by the large increase in tangent modulus in Figure 6 (2.2–7.7-fold increase from 10 to 45 mm Hg in the peripapillary sclera from normal eyes), and also by the large increase in structural stiffness as IOP increased (Fig. 7). This phenomenon is also demonstrated by the nonlinear relationship between maximum principal strain and IOP (Fig. 8). Note that maximum principal strain values were consistently low in the normal eyes, as shown in blue in Figure 8, and varied between 0.32% and 1.49% on average in the peripapillary sclera at 45 mm Hg. Conversely, maximum principal stress increased linearly with IOP (Fig. 8) and was considerably higher than IOP (16 to 30 × IOP on average in the peripapillary sclera from normal eyes at all IOP levels).
For each eye, the sclera was highly inhomogeneous as demonstrated by the regional variations in tangent modulus (Fig. 4, row 4). On average, the tangent modulus was higher in the peripheral sclera (17–60 MPa in the normal eyes) than in the peripapillary sclera (15–50 MPa in the normal eyes).
The sclera was highly anisotropic in all eyes as demonstrated by the non-zero fiber concentration factors shown in Table 1 and Figure 2. The fiber concentration factor in the peripapillary sclera (k2) was higher than that in the peripheral sclera (k1) in 10 of the 16 eyes studied (Table 1). Maps of the predicted preferred fiber orientation for all eight regions of each eye are shown in Figure 4, where (white) indicates a preferred fiber orientation perpendicular to the scleral canal (meridional, θp = 0°), and // (black) indicates a preferred fiber orientation tangent to the scleral canal (circumferential, θp = 90°). Figure 9 shows the pooled distributions of the preferred fiber orientation for all 16 eyes in both the peripapillary and peripheral sclera. In general, the mean preferred fiber orientation for all tested eyes was tangent to the scleral canal in both scleral regions, but this effect was more pronounced in the peripapillary sclera.
We considered all eight normal eyes from this study as being part of one age group (adult, 17.3 ± 5.1 years). Note that the monkey lifespan is about one third the human lifespan.51 We compared tangent modulus and structural stiffness distributions from the monkeys in this study to that of two other age groups (young, 1.5 ± 0.7 years; old, 22.9 ± 5.3 years) from our previous report.8 Using the generalized estimating equation method, we found that the tangent modulus and structural stiffness in the peripapillary sclera significantly increased with age (Table 2). Using a multiple linear regression, we found that eyes from old monkeys were significantly thinner than eyes from adult monkeys (P < 0.05) but found no significant differences in thickness between eyes of young and adult monkeys (P > 0.05). Peripapillary scleral thickness averaged 389 ± 53 μm in young eyes,8 381 ± 47 μm in adult eyes (this study), and 349 ± 49 μm in old eyes.8
We observed no significant changes (exceeding the physiologic intereye difference of normal monkeys) in tangent modulus (Fig. 6) and structural stiffness (Fig. 7) between the normal and glaucomatous eyes of monkeys 2, 3, and 8.
We observed significant changes in tangent modulus (Fig. 6) and structural stiffness (Fig. 7) between the normal and glaucomatous eyes of monkeys 1, 4, 5, 6, and 7. The glaucomatous eye in monkey 1 exhibited a significant decrease in tangent modulus (10 mm Hg) and structural stiffness (10 and 30 mm Hg). Note that monkey 1 received the lowest cumulative IOP insult and these changes manifest at lower IOPs. The glaucomatous eyes in monkeys 4, 5, and 6 exhibited a significant increase in tangent modulus and structural stiffness at higher IOPs (30 and 45 mm Hg). Monkey 7 exhibited a significant increase in tangent modulus at 45 mm Hg in the peripheral sclera only.
We observed significant changes (exceeding the physiologic intereye difference of normal monkeys) in maximum principal strain in two of the eight monkeys (Fig. 8, 1 and 4) and in maximum principal stress in two of the eight monkeys (Fig. 8; 4 and 6).
We did not detect any changes in the fiber concentration factor (P > 0.05) or in the preferred fiber orientation (P > 0.05) in either the peripapillary or peripheral regions between the normal and glaucomatous eyes.
Finally, peripapillary and peripheral scleral thickness in both the normal and glaucomatous eyes of each monkey are shown in Figure 5. Although a trend toward scleral thinning was observed in the glaucomatous monkey eyes, overall and regional change in scleral thickness between glaucomatous and normal eyes was insignificant when compared to the physiologic intereye difference from eight normal monkey eyes (P > 0.05).
In this study, we characterized scleral biomechanical behavior in both eyes of eight adult monkeys in which one eye had been exposed to chronic, laser-induced IOP elevations of modest to substantial magnitude and duration. The following are the principal findings of this report.
When considered as a group of 16 monkey eyes without regard for experimental condition (normal versus EG) the eyes in this report behaved similarly to the eyes from bilaterally normal monkeys from a previous report,8 as follows. First, the monkey posterior sclera is thicker in the peripapillary and temporal regions. Second, the monkey posterior sclera is a nonlinear, anisotropic, and inhomogeneous soft tissue. Third, concentrations of tangent modulus, structural stiffness, and maximum principal stress and strain were evident adjacent to the scleral canal in each eye. Fourth, maximum principal stress magnitude was substantially higher than IOP. Fifth, maximum principal strain magnitude was low. A detailed discussion of these results can be found in our previous report.8
Scleral biomechanical properties (i.e., tangent modulus and structural stiffness; Table 2) in the eight normal adult eyes (mean age, 17.3 ± 5.1 years) from this study fell between those from young (1.5 ± 0.7 years) and old (22.9 ± 5.3 years) monkeys from our previous aging study.8 These findings separately confirm our initial report of age-related stiffening of the posterior sclera by establishing that adult eyes are stiffer than young ones but less stiff than old eyes. This result is likely to be related to increasing scleral cross-links with age28,52; details on this topic can be found in our previous report.8
We have reported a decrease in scleral thickness in old compared with young normal monkey eyes. In that study, we found that adult monkey sclera was not significantly thinner than young monkey sclera, but was significantly thicker than sclera from old monkey eyes. This finding is important because it suggests that the rate of scleral thinning, if present, may accelerate with age. Whether scleral thinning with age could accelerate the progression of glaucoma in elderly patients remains unclear. In monkeys, we observed a reduction in peripapillary scleral thickness of approximately 8.4% from adult to older eyes on average. In comparison, we observed an increase in tangent modulus of 37.4% (at 30 mm Hg). If both tangent modulus and thickness contribute similarly to the overall stiffness of the scleral shell (i.e., structural stiffness), then age-related changes in peripapillary scleral thickness will have a lesser impact on IOP-induced deformation than those of the tangent modulus. The relationship between glaucomatous susceptibility and scleral stiffness is likely to be complex, individual specific, and involve many interacting factors.
The major findings were as follows: First, significant changes in tangent modulus and structural stiffness were observed in some monkeys between the normal eyes and those exposed to chronic IOP elevation. Second, changes in FE-based scleral anisotropy measures were not observed between normal and glaucomatous monkey eyes. Finally, although all glaucomatous eyes exhibited thinner peripapillary sclera than their contralateral normal eyes (Fig. 5), these differences did not achieve statistical significance.
We observed significant changes in scleral tangent modulus (Fig. 6) and structural stiffness (Fig. 7) between the normal and glaucomatous eyes of monkeys 1, 4, 5, 6, and 7. The glaucomatous eye in monkey 1 exhibited a smaller tangent modulus at 10 mm Hg (Fig. 6) and structural stiffness at 10 and 30 mm Hg (Fig. 7) than those of its contralateral normal. Burgoyne et al.53 have reported hypercompliance of the ONH surface within 4 to 8 weeks of chronic IOP elevation in monkeys and suggested that weakening of the load-bearing structures of the eye (lamina cribrosa and peripapillary sclera) through connective tissue damage might be responsible for this phenomenon at the earliest stage of glaucomatous optic neuropathy. Moreover, a computational modeling study by Downs et al.54 reported weakening of the lamina cribrosa structure at the earliest stage of glaucomatous optic neuropathy. It is interesting to note that weakening of the sclera was only present in the glaucomatous eye of monkey 1, the eye that received the smallest cumulative IOP insult (432 mm Hg × days) and had the lowest maximum measured IOP (24 mm Hg). It should also be noted that the decrease in tangent modulus in the glaucomatous eye of monkey 1 was only significant at low IOP (10 mm Hg).
In 1957, Roach and Burton55 studied the roles of elastin and collagen in arteries. They found that elastin alone was responsible for resisting pressure-induced arterial deformations at low pressure and that collagen alone was responsible for resisting pressure-induced arterial deformations at high pressure. These results were recently confirmed by Fonck et al.56 Based on these studies, it could be that only damage to elastin occurred in the glaucomatous eye of monkey 1, and this represents the earliest stage of scleral damage and/or remodeling in glaucomatous progression.
We observed a significant increase in scleral tangent modulus at IOPs of 30 and 45 mm Hg in both the peripapillary and peripheral regions in monkeys 4, 5, and 6, and a significant increase in scleral tangent modulus at 45 mm Hg in the peripapillary region in monkey 7 (Fig. 6). The glaucomatous eyes in monkeys 4, 5, and 6 also exhibited larger structural stiffness in the peripapillary sclera at IOPs of 30 and 45 mm Hg (Fig. 7). These findings suggest that at higher levels of cumulative IOP insult, the posterior sclera stiffens through extracellular matrix remodeling. This finding is consistent with that in our previous uniaxial study of scleral biomechanics in glaucomatous monkey eyes11 and may be a protective mechanism that limits scleral canal expansion and related deformation of the ONH at elevated IOPs. Zeimer and Ogura57 reported that human ONHs became more rigid with severe glaucoma, a result that is consistent with this hypothesis. Roach and Burton55 also demonstrated that collagen alone was responsible for limiting arterial deformations at high pressure, which suggests that the observed increase in tangent modulus in glaucomatous eyes at high IOPs may be due primarily to collagen remodeling.
Even though tangent modulus changes occurred in five monkeys (1, 4, 5, 6, and 7), significant changes in maximum principal strains (monkeys 1 and 4) and maximum principal stress (monkeys 4 and 6, in the peripapillary sclera only) were each observed in only two monkeys. This result suggests that overall changes in scleral thickness (generally lower in the glaucoma eyes) and tangent modulus (generally higher in the glaucoma eyes) at this stage of damage interact to maintain the stress and strain ranges typical of sclera in normal eyes.
In this study, we observed no significant changes in either scleral tangent modulus (Fig. 6) or structural stiffness (Fig. 7) between the normal and glaucomatous eyes in three monkeys (2, 3, and 8). Since monkey 8 received the highest cumulative IOP insult (7714 mm Hg × days), we would have expected to observe the largest changes in the glaucomatous eye of this monkey. Of note, at IOPs of 30 and 45 mm Hg, the normal eyes of monkeys 3 and 8 had the two highest mean tangent moduli and structural stiffnesses in the peripapillary sclera. Assuming that before IOP insult, the glaucomatous eyes in monkeys 3 and 8 were as stiff as their contralateral normals, these results suggest that stiffer scleral shells may be less prone to biomechanical changes when exposed to chronic IOP elevation, presumably because they are more resistant to IOP-induced strain.
Several studies have reported that mechanical strain applied to scleral fibroblasts can trigger the release of MMPs and TIMPs,58–61 which can then lead to remodeling of the scleral extracellular matrix. IOP-induced scleral shell expansion increases scleral strain in all eyes, but in eyes with the stiffest sclera, these increases may be too small to elicit a fibroblast response. Note that mean strain levels at 45 mm Hg in monkeys 3 and 8 were 0.30% and 0.32% respectively, which were smaller than all the reported values in the aforementioned in vitro studies on scleral fibroblast response to strain (a minimum of 0.45%58).
In the case of the third monkey, which demonstrated no change in tangent modulus or structural stiffness (monkey 2), the normal eye had the most scleral thickness (Fig. 5), and lowest tangent modulus (Fig. 6) among all normal eyes. Its mean structural stiffness was therefore comparable with the mean structural stiffness of all normal eyes. These data suggest that mean structural stiffness may not always predict which eyes will undergo IOP-induced scleral remodeling and that scleral thickness may also be an important contributor in predicting which eyes will be less sensitive to biomechanical change.5,46
In this study, we did not observe significant differences in predicted scleral anisotropy (preferred fiber orientation and fiber concentration factor) between normal and glaucomatous monkey eyes. The preservation of anisotropy was recently reported for the remodeled monkey lamina cribrosa in EG,15 suggesting that remodeling of the peripapillary and posterior sclera must be accompanied by changes in elastin/collagen content and/or fiber diameter, but not fiber reorientation. This is consistent with the work of Quigley et al.,62 who reported glaucomatous changes in collagen and elastin content and collagen fiber diameter in the peripapillary sclera in human and monkey eyes.
Although we did observe a trend toward scleral thinning in the glaucomatous eyes, this difference did not exceed the physiologic intereye differences in the eight normal monkeys. Using traditional histology, we previously reported significant thinning of the posterior10 and peripapillary sclera13,14 in monkey eyes subjected to chronic IOP elevation. Scleral morphologic changes should be assessed in future studies to understand whether scleral thinning due to chronic IOP elevation is a consequence of strain-induced fluid exudation, tissue loss, or a combination of the two.
Overall, our data suggest that a stiffer scleral shell is biomechanically beneficial and is less apt to remodel in response to elevated IOP. In normal eyes, we suggest that the sclera shields the ONH from biomechanical insult by resisting large overall deformation (through uncrimping and stiffening of the collagen fibers: nonlinearity) and large scleral canal expansion (through the presence of a circumferential peripapillary ring of collagen fibers: anisotropy).3,4,8 We hypothesize that the scleral stiffening in response to moderate exposure to chronic IOP elevations shown herein is an effort to maintain the biomechanical homeostasis of the scleral shell and indirectly, the ONH. As such, this response could act to slow glaucomatous progression in eyes with higher IOPs and/or more compliant scleral shells. Surprisingly, the argument that a stiff scleral shell shields the ONH from glaucomatous damage seems to contradict the facts that (1) the sclera stiffens with age.8 and (2) the elderly are more susceptible to glaucoma.63 To reconcile all these observations, it should be noted that collagen fibers become brittle with age28 and therefore it is plausible to suggest that scleral stiffening with age is an unhealthy mechanism as opposed to scleral stiffening because of the remodeling seen in this study. In addition, there are other factors that are likely to contribute to glaucomatous damage, including blood flow and cellular reactivity, and these factors change with age as well. We also caution the reader that stiffness is a broad term, and reporting a single stiffness value for the sclera does not represent its biomechanical response well. Scleral stiffness is a function of IOP (nonlinearity), orientation (anisotropy), and location (heterogeneity). This complexity should be taken into account when evaluating the contribution of scleral biomechanics to glaucoma pathogenesis.
Several limitations should be recognized when viewing this work. Although the general limitations of the method have been discussed at length in our previous reports,3,8,9 we will briefly revisit them and then discuss those inherent to this study.
First, IOP was measured in both eyes of each monkey over time (Fig. 1), but we did not characterize the diurnal and nocturnal IOP fluctuations that are known to exist in monkeys. Diurnal IOP fluctuations have been shown to be large in monkeys, especially in glaucomatous eyes, in which they are highly individual-specific and can exceed 10 mm Hg.64 Therefore, our definition of cumulative IOP insult (change in area under the IOP/time curve) should be regarded with caution and was used here only to provide a rough estimate of the cumulative insult from IOP. An implantable telemetric IOP transducer is currently being developed for monkeys that should allow for continuous measurements of diurnal and nocturnal IOP fluctuations and more accurate characterization of IOP insult (Downs JC, et al. IOVS 2008;49:ARVO E-Abstract 2043).
Second, because the sclera is approximately 90% collagen by weight,65 we assumed that collagen is the primary fibrous element66 of this tissue. We lumped all other tissue constituents into the ground substance matrix, a common practice in soft-tissue FE modeling.4,30,67–71 Because our study suggests that both elastin and collagen play important roles in glaucomatous damage, further experimental and theoretical work is needed to address how these components separately contribute to scleral biomechanics.
Third, we did not characterize scleral biomechanics between 0 and 5 mm Hg in the tested eyes, because the scleral shells required an IOP of approximately 4 mm Hg to sustain their shape. IOPs in the range of 0 to 5 mm Hg are rarely measured in monkey eyes, and so ignoring the initial IOP loading should not compromise the clinical importance of our results. We are developing methods that will allow the initial scleral stress and strain that exist in the tissues at 5 mm Hg to be estimated and included in future studies.
Fourth, the ONH tissues were assumed to be linearly elastic, with a common elastic modulus of 1 MPa assigned to the ONH region for both normal and glaucomatous monkey eyes. Our previous sensitivity study showed that ONH elastic moduli in the range of 0.1 to 5 MPa did not affect the results of our scleral biomechanical property fitting9; hence, our assumption should not affect the results presented herein.
Fifth, within our models the ONH was represented as an elliptical cylinder, with its external boundary located at the outer aspect of the dural sheath insertion (as measured with the 3-D digitizer) and therefore included regions of the immediate peripapillary sclera and scleral flange. As a result of this anatomic convention, we could not estimate biomechanical properties for that region of the sclera that lies immediately under the dural sheath insertion.
Sixth, we divided each reconstructed scleral shell into eight regions to assess regional variations in predicted collagen fiber orientation (θp1–θp8). This division was made because a preliminary study confirmed that a combination of eight regions and 13 model parameters was the minimum requirement to obtain good agreement between the FE-predicted and experimentally measured displacements (Fig. 4), while also ensuring a unique set of model parameters. Future studies will include a more rigorous assessment of regional variability.
Seventh, all experiments were conducted at room temperature (22°C) because the incorporation of a heated saline chamber at 37°C proved incompatible with the speckle interferometry sensor we used. Because all experiments were consistently performed at the same temperature, all hypotheses tested regarding the changes in scleral biomechanics should be valid.
Eighth, collagen fiber alignment was not measured directly, but was predicted using the models' fits to experimental scleral deformations (Fig. 4). Because of the relatively coarse discretization of the regions (discussed above) we could only report an overall approximation of the preferred collagen fiber orientation within each region. Future inclusion of experimental measures of collagen fiber orientation as input parameters, as currently being measured with small-angle light-scattering (Girard MJ, et al. IOVS 2010;51:ARVO E-Abstract 2128), will allow for finer regional discretization for the remaining fitted parameters and improve our characterization of scleral anisotropy.
Finally, seven of the eight monkeys (monkeys 2–8) began this study while the investigators laboratories were still located at the LSU Eye Center in New Orleans, and were unavailable for study for more than 3 months after Hurricane Katrina. Monkeys 2 and 3 and 6 to 8, however, had already undergone initial laser, and so the accuracy of their cumulative IOP exposure calculations is affected by the lack of IOP data between 30 August and 15 December 2005, (Fig. 1, shaded area) and the diminished frequency of testing after their arrival at Devers. In addition, these monkeys (2, 3, and 6 to 8) were examined only sporadically until the laboratory at Devers became fully operational. At the time of the hurricane, monkeys 4 and 5 had not yet been lasered, and so the frequency of IOP measurement and assessment of pre- and postlaser cumulative IOP exposure in these animals was unaffected by the hurricane. Monkey 1 was purchased at Devers and was altogether unaffected by these events.
In summary, new experimental and computational methodologies3,9 were used in this study to characterize scleral biomechanics in normal and glaucomatous monkey eyes. These methods have broad applicability and can be applied to other thin, soft tissues with multidirectional collagen fibers. The long-term goal of this work is to establish the pathophysiologic mechanisms linking connective tissue stress and strain within the peripapillary sclera and lamina cribrosa to retinal ganglion cell axon damage and death. Given the important role of scleral biomechanics in determining the biomechanics of the ONH, new strategies for the clinical visualization and measurement of peripapillary and posterior scleral biomechanics are needed.
Overall, our results suggest that significant changes in the biomechanical behavior of the sclera are associated with exposure to chronic IOP elevations. These changes are complex and individual-specific and are likely to be the result of scleral extracellular matrix remodeling. Our findings suggest that (1) scleras with initially large tangent modulus or thickness are more resistant to IOP-induced biomechanical changes, (2) a decrease in tangent modulus (possibly due to elastin damage) occurs at the earliest stage of glaucoma, (3) an increase in tangent modulus (possibly associated with collagen changes) occurs at moderate stages of glaucoma, and (4) collagen fiber reorientation does not occur on a large scale in glaucomatous eyes. The sclera drives the levels of stress and strain transmitted to the ONH, so changes in scleral biomechanical behavior will inevitably perturb the biomechanics of the ONH.7 Scleral remodeling induced by exposure to chronic IOP elevations may provide a protective mechanism for the contained ONH and act to keep stress and strain levels within the normal range, but future work is necessary to establish the link between scleral/ONH biomechanics and glaucomatous vision loss.72–74
The authors thank Juan Reynaud, Jonathan Grimm, Wenxia Wang, and Stuart Gardiner for their contributions to this work.
Presented in part at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, May 2009.
Supported by NIH Grant R01EY11610 (CFB) and the Legacy Good Samaritan Foundation, Portland, OR.
Disclosure: M.J.A. Girard, None; J.-K.F. Suh, None; M. Bottlang, None; C.F. Burgoyne, None; J.C. Downs, None