|Home | About | Journals | Submit | Contact Us | Français|
To characterize optic nerve head (ONH) connective tissue deformation following acute (15 or 30 minutes) intraocular pressure (IOP) elevation within six adult normal monkeys using 3-D histomorphometry.
Trephinated ONH and peripapillary sclera from both eyes of six monkeys, each perfusion fixed with one eye at IOP 10 mmHg and the other at IOP 30 or 45 mmHg by anterior chamber manometer were serial sectioned, 3-D reconstructed, 3-D delineated and quantified using standard parameters. For each monkey, inter-eye differences (high IOP eye minus IOP 10 eye) for each parameter were calculated and compared by ANOVA and EPIDmax both overall and regionally. EPIDmax deformations for each parameter were defined to be those statistically significant differences that exceeded the maximum physiologic inter-eye difference within six bilaterally normal monkeys of a previous report.
Regional EPIDmax laminar thinning, posterior bowing of the peripapillary sclera, thinning and expansion of the scleral canal were present in most high IOP eyes and were colocalized in those demonstrating the most deformation. Laminar deformation was minimal and not only posterior but in some cases anterior in the high IOP eyes. No increase in deformation was seen in the IOP-45 versus the IOP-30 eyes.
ONH connective tissue alterations following acute IOP elevation involve regional thinning, stretching and deformation of the lamina cribrosa and peripapillary sclera which are minimal to modest in magnitude. The time-dependent character of these alterations, as well as their compressive, expansile, and shear effects on the contained axons, astrocytes, laminar and posterior ciliary circulations remain to be determined.
The neural, vascular, and connective tissues of the optic nerve head (ONH) make up a dynamic environment wherein 1.2 to 2.0 million retinal ganglion cell (RGC) axons converge, turn, and exit the eye through the neural canal opening (NCO; Figure 2 and Figure 7). Within the scleral portion of the neural canal, the bundled axons pass through the lamina cribrosa, a three-dimensional (3-D) meshwork of astrocyte-covered, capillary-containing, connective tissue beams. While glaucomatous damage to the visual system likely includes important pathophysiologies within the RGC body 1-6, photoreceptors7-11, lateral geniculate body12-14 and visual cortex14, strong evidence suggests that damage to the retinal ganglion cell axons within the lamina cribrosa of the ONH15-20 is the central pathophysiology underlying glaucomatous vision loss. Recent studies in human21, 22, monkey19, 20, 23-28, rat29-31 and mouse32, support the importance of the ONH by describing significant alterations within the prelaminar, laminar and peripapillary scleral tissue of the ONH at the earliest detectable stage of experimental glaucoma. While the interaction between ONH connective tissue deformation19, 20, 24, 26, 33, altered astrocyte function34-38 and ischemia 39-42 are beginning to be elucidated, the multifactorial insults to the RGC axons remain to be characterized. These insults may include intraocular pressure (IOP)-related27 and non IOP-related components43-46 that do not require IOP-induced connective tissue deformation, damage, and/or remodeling to occur.
To characterize the ONH as a biomechanical structure, deformation of the lamina cribrosa following acute IOP elevations has been studied in human cadaver eyes21, 22, 47 and in monkeys19, 48. Levy and Crapps22 reported a 12 μm average posterior (outward) movement of the central lamina with acute IOP elevations from 10 to 25 mmHg for short periods in human cadaver eyes. Yan et al.21 found that increasing IOP from 5 to 50 mmHg for 24 hours produced an average posterior deformation of the central lamina of 79 μm, and small contractions of the scleral canal and thinning of the lamina which did not achieve significance.
We have previously studied acute IOP-induced deformation of the monkey ONH connective tissues using 2-D histomorphometry within two groups of monkey eyes 19, 48. In the first study 48, Bellezza demonstrated that anterior laminar position was significant more anterior (towards Bruch's membrane), the lamina cribrosa was thinner, and the scleral canal diameter was larger in a group of monkey eyes that had been perfusion fixed at IOP-10 mmHg compared with a group of immersion fixed (IOP-0 mmHg) eyes. These results suggested that the lamina cribrosa and scleral canal wall act like an expandable trampoline at low levels of IOP, with the canal expanding and the lamina thinning and becoming more tautly stretched as IOP is elevated from 0 to 10 mmHg.
In a subsequent study 19, young adult normal monkeys were perfusion fixed 15 to 80 minutes after one eye was set to 10 mmHg and the other to 30 or 45 mmHg by anterior chamber manometer. This study demonstrated an overall posterior laminar deformation of 10 to 23 μm in the high IOP eyes relative to their contralateral IOP-10 control eyes. However, this deformation did not exceed the magnitude of inter-eye differences in laminar surface position in a separate group of bilaterally normal monkeys immersion fixed following enucleation.
These human and monkey studies, considered together, suggest that the lamina cribrosa and scleral canal wall deform following acute IOP elevation in the normal monkey and human eye, that the deformation of the lamina may be anterior rather than posterior within the canal, and that this deformation may increase as the period of acute IOP elevation is extended. However, the measurements in most of these studies were taken from two dimensional (2-D) histologic sections and were therefore subject to artifacts introduced by the section angle through the tissue, section distortion during mounting, and the lack of a consistent reference plane for measurements, all of which may have limited their accuracy. Levy's study using 2-D x-ray images also suffered from artifacts induced by the angle of imaging and the fact that the position of the platinum wire was not certain relative to the anterior lamina cribrosa surface.
While we have previously emphasized that IOP-related axonal trauma within the lamina cribrosa may be unrelated to actual connective tissue deformation25, 33, it remains important to identify the biomechanical determinants of laminar and peripapillary scleral displacement at a given IOP. The purpose of this study is to characterize connective tissue deformation of the lamina, scleral canal wall, and peripapillary sclera within 3-D ONH reconstructions following acute (15 minutes) IOP elevation (from 10 to 30 or 45 mmHg) within six young adult normal monkeys using 3-D histomorphometry24-26, 28
All animals were treated in accordance with the Association for Research in Vision and Ophthalmology (ARVO) Statement for the Use of Animals in Ophthalmic and Vision Research. One female and five male normal monkeys (5 - 14 years old), were used for this study (Table 1). Prior to sacrifice, IOP and axial length were measured in both eyes of each animal on 3 to 5 separate occasions. IOP was defined to be the mean of three measurements by Tono-pen (Tono-Pen XL; Bio-Rad, Glendale, CA). Axial length was defined to be the mean of the longest 3 measurements out of 10 ultrasonic axial length measurements (Model A1500; Sonomed, Lake Success, NY).
Both eyes of each monkey were cannulated with a 27-gauge needle under deep pentobarbital anesthesia and the IOP was set to 10 mmHg using an adjustable saline reservoir. After a minimum of 30 minutes, IOP was raised to either 30 or 45 mmHg in one eye for 15 minutes (30 minutes for monkey 6) and the monkey was perfusion fixed via the descending aorta with 1 L of 4% buffered hypertonic paraformaldehyde solution followed by 6 L of 5% buffered hypertonic glutaraldehyde solution pressurized during perfusion to approximately 4 psi for 5 monkeys, and 15 psi for one monkey (monkey 1 – the only monkey in which perfusion pressure was monitored via the brachial artery during perfusion). Following perfusion, IOP was maintained for one hour, and then each eye was enucleated, all extraocular tissues were removed, and the intact anterior chamber was excised 2 to 3 mm posterior to the limbus. By gross inspection, perfusion was excellent in all six IOP-10 mmHg eyes. However, blood was variably present in the retinal vessels, posterior ciliary arteries and vortex veins of the six high IOP eyes. The posterior scleral shell with intact ONH, choroid and retina were placed in 5% glutaraldehyde solution for storage.
These steps have been described in detail within our previous reports20, 24. Briefly, the ONH and peripapillary sclera were trephinated (6-mm-diameter), pierced with alignment sutures, embedded in paraffin, mounted to a microtome and sectioned. After each section was taken, the block surface was stained with a 1:1 (v/v) mixture of Ponceau S and acid fuchsin stains, and then imaged at a resolution of 2.5 × 2.5 μm per pixel. Sections were serially cut at 3.0 μm thickness, and the staining and imaging process was repeated after each cut. Imaging began at the vitreoretinal interface and continued approximately 200 μm into the retrolaminar orbital optic nerve.
Serial section images of ONH were aligned in the anterior-to-posterior direction and stacked at 3.0-μm intervals into a 3-D reconstruction of the ONH and peripapillary scleral connective tissue (approximately 1080 × 1520 × 400 voxels, each 2.5 × 2.5 × 3.0 μm in size).
Our 3-D delineation technique has been described in detail in previous reports24-26 (see Figure 1). Briefly, using custom software (based on the Visualization Toolkit, Clifton Park, NY), the 3-D ONH reconstruction was loaded and the delineator designated the approximate center of the neural canal as the center of rotation, around which 40, 7-voxel-thick, digital, radial, sagittal slices of the 3-D reconstruction were serially served at 4.5° intervals to the delineator's workstation (Figure 1A).
Within each sagittal section image, the delineator marked seven structures: the lamina cribrosa, sclera, neural boundary, Bruch's membrane, internal limiting membrane, central retinal vessels, and subarachnoid space; and six pairs of neural canal landmark points: the Neural Canal Opening (NCO) which in the monkey is most often the innermost extension of Bruch's membrane opening (BMO), the anterior scleral canal opening (ASCO), the anterior laminar insertion (ALI), the posterior laminar insertion (PLI), the posterior scleral canal opening (PSCO), and the anterior end of the subarachnoid space (ASAS) (Figure 1B). While marking in the sagittal section view window, the delineator simultaneously viewed a slaved window showing the cursor's location within a digital transverse section image (Figure 1C). The 3-D Cartesian coordinates and category for each mark were saved, generating a 3-D point cloud that represented each of the marked structures (Figure 1D).
Three experienced delineators performed all of the delineations in this study. Both eyes of each animal were delineated by a single delineator. Delineators were not masked to the IOP status of each eye, however, they were effectively unaware of the IOP status during delineation because the file name of each volume within the delineation software did not contain that information. Upon completion of delineation of both eyes of each monkey, the marks were checked for accuracy by two authors with extensive experience (HY and CFB) and without knowledge of the IOP status.
The inter- and intra-delineator variability and reproducibility of our delineation technique has previously been characterized in a series of reports 24-26. Briefly, inter-delineator variability for each parameter was assessed by having five delineators (only 2 delineators for our cupping parameters 25) delineate all ONH landmark points within the normal and early glaucoma eyes of three monkeys. Intra-delineator reproducibility was assessed by having two of the five delineators (only one delineator for our cupping parameters) delineate all landmarks of both eyes of one monkey on two additional occasions at least 2 weeks apart. The inter- and intra-delineator variability data from the three pairs of monkeys (one eye is normal and the contralateral eye is early-glaucoma) of our previous reports are summarized in Table 2.
For each ONH, a reconstruction of the central retinal vessels was performed and three-dimensionally overlaid onto a pre-sacrifice photo using the qualitative match of the ONH and retinal vessels. Once preliminarily aligned (using the vessels only), the vessels and NCO points were co-visualized to assess the relationship of the optic disc margin in the photo to the delineated NCO points. A final 3-D adjustment was then performed to match NCO to the disc margin while maintaining vessel alignment.
NCO Reference Plane For each 3-D ONH reconstruction, a least-squares ellipse was fit to the 80 NCO points (which in this case are the end of Bruch's membrane) 24, 49 creating an NCO reference plane (Figure 2C)24. The centroid of the NCO ellipse established the center point for all measurements. All quantification of offset, depth, position, and post-NCO total prelaminar volume were made relative to this plane (Figure 2C-2G).
Quantification This report includes overall and regional quantification of neural canal offset and depth, lamina cribrosa position and thickness, scleral flange thickness, peripapillary scleral position and thickness, and post-NCO total prelaminar volume. The definitions and calculation methods of these parameters have been described in detail in our previous reports24-26, and are briefly summarized as follows (Figure 2). Neural canal landmark Depth is the anterior-to-posterior distance of each marked point to the NCO zero reference plane; and Offset is the distance of each mark, projected onto the NCO reference plane, from the NCO centroid. Laminar Position is the shortest distance from each delineated anterior laminar surface point to the NCO reference plane. Laminar Thickness was calculated at each delineated anterior laminar surface point as the shortest distance to the posterior laminar surface along a vector normal to the anterior laminar surface (fitted into continuous surface to generate the normal vector).
Peripapillary Scleral Position and Thickness was calculated in the same way as laminar position and thickness. Scleral Flange Thickness is defined as the distance from the anterior scleral flange surface to the neural canal boundary surface, measured at each delineated anterior scleral flange surface point along a vector parallel to the PSCO normal vector. Post-NCO Total Prelaminar Volume is designed to detect deformation of the anterior neural canal connective tissue in a single parameter, and is defined to be the total volume (tissue and non-tissue) beneath the NCO reference plane, above the lamina cribrosa and within the neural canal wall.
Overall and regional parameter values were calculated for both eyes of each monkey as outlined in Figure 3.
Schematic depictions of overall (Figure 4) and regional (Figure 5) differences were created without regard for statistical significance for each monkey by overlaying the overall mean measurements for each eye (Figure 4) or for the superior and inferior regions (left Figure 5) and nasal temporal regions (right Figure 5) of each eye.
For each ONH, 80 measurements of the following parameters were made: offset and depth of the ASCO, ALI, PLI, PSCO, and ASAS. Measurement of laminar position and thickness, scleral flange thickness, and peripapillary scleral position and thickness were made at each delineation point after surfaces were fit to the anterior and posterior laminar and scleral points. Therefore for all non-volumetric parameters at least n=40 measurements were made for each ONH, and a factorial analysis of variance (ANOVA) was performed to assess the effect of region and IOP (IOP 10 vs IOP 30 or IOP 10 vs IOP 45) on each parameter between the two eyes of each monkey. Statistically significant differences between regions, IOPs, and region-by-IOP combinations required an overall significant F test followed by t tests using p values corrected for multiple comparisons50. Because our volumetric parameters do not lend themselves to multiple measures within individual eyes, a statistical assessment of intra-eye differences for Post-NCO total prelaminar volume for each monkey was not possible.
Statistically significant differences were determined by an ANOVA as outlined above. We additionally defined EPIDmax differences to be those statistically significant differences that exceeded the physiologic inter-eye difference (PID) maximum (the maximum inter-eye difference) within 6 bilaterally normal monkeys in a previous report28. Within this construct, we considered statistically significant differences between the high and low IOP eye of each monkey to be those that exceeded the combination of the variability of our 3-D Histomorphometric quantification method and the true biological variation of each parameter within a region. We considered EPIDmax differences to be those statistically significant differences that were more likely to be treatment effects, i.e. represent a biological response to acute pressure elevation rather than the difference between the two eyes of a normal monkey. For each volumetric parameter, we compared the inter-eye difference for each animal to the PID maximum difference for that parameter.
We measured post-mortem inter-eye differences for each parameter for each monkey. We emphasize EPIDmax inter-eye differences as those statistically significant differences that most likely represent acute IOP induced change in the high IOP eye of each monkey. We thus refer to EPIDmax inter-eye differences as “change” or “EPIDmax change”.
We further grouped our parameters into “position-change” and “thickness-change” parameters to better recognize the colocalized behavior of these two forms of deformation. To describe position-change we employed the following terminology as it relates to EPIDmax differences in our measurement parameters: Neural canal radial expansion was defined to be an EPIDmax increase, neural canal radial contraction an EPIDmax decrease in neural canal offset, lamina cribrosa anterior deformation was defined to be an EPIDmax increase (more anterior or inward), and lamina cribrosa posterior deformation an EPIDmax decrease (more posterior or outward) in anterior lamina cribrosa position relative to the NCO reference plane.
Peripapillary scleral posterior bowing was defined to be a posterior (outward) deformation of the peripapillary sclera relative to the more peripheral sclera which manifests as an EPIDmax increase in the position of the peripapillary sclera. In this scenario, because the peripapillary sclera carries the ONH with it as it moves outward, it assumes a final position that is anterior to the NCO reference plane. Post-NCO total prelaminar volume expansion was defined to be an EPIDmax volumetric increase andpost-NCO total prelaminar volume contraction an EPIDmax volumetric decrease in the space confined below the NCO reference plane, within the neural canal wall and above the anterior laminar surface. As such, this parameter volumetrically combines the individual deformations of the neural canal wall and lamina cribrosa relative to the NCO reference plane.
To describe thickness-change, we defined neural canal axial thickening to be an EPIDmax increase, and neural canal axial thinning an EPIDmax decrease in neural canal landmark depth. We defined laminar and peripapillary scleral thickening and thinning as EPIDmax increases or decreases in the thickness measurements for each tissue.
Descriptive data for the six normal monkeys and the histomorphometric optic disc size for each eye are reported in Table 1. Five rhesus monkeys and one cynomolgus monkey, aged from 5 to 14 years, were used for this study. Mean IOP under ketamine/xylazine anesthesia varied from 4 to 12 mmHg among both eyes of all six monkeys. Ultrasonic axial length measurements prior to sacrifice (Model A1500; Sonomed, Lake Success, NY) ranged from 18.29 - 21.39 mm.
Histomorphometric vertical disc size (measured at NCO) ranged from 1214 μm to 1641 μm, and histomorphometric horizontal disc size ranged from 864 μm to 1143 μm. Histomorphometric optic disc area ranged from 0.801 to 1.473 mm2 (based on an ellipse fit to 80 delineated NCO points – see methods).
Overall IOP-10 eye data along with the statistically significant and EPIDmax differences for the high IOP eye for each animal are reported in Table 2. Schematic plots of the overall data for both eyes of each monkey are presented in Figure 4.
Within the schematic plots of Figure 4, qualitative differences between the 6 normal eyes include laminar surface curvature (relatively flat in monkey 5 and relatively curved in monkeys 1 and 4); laminar and peripapillary thickness (relatively thin in monkey 1 and relatively thick in monkey 6); and neural canal opening size and obliqueness (small and less oblique in monkey 6, and large and more oblique monkey 3). Overall, the lamina cribrosa did not deform appreciably in the high IOP eyes of any of these monkeys compared to their contralateral IOP-10 eye. Inter-eye differences following acute IOP elevation are greatest in monkeys 5 and 6 and include expansion and thinning of the posterior neural canal, accompanied by thinning of the lamina without substantial posterior deformation and posterior bowing of the peripapillary sclera.
Within the overall data in Table 2, very little change achieved EPIDmax differences and therefore we emphasize the regional (below) rather than the overall results.
Superimposed, averaged, central vertical and horizontal data for both eyes of each animal are schematically depicted in Figure 5. Statistically significant and EPIDmax regional differences for each parameter and for each monkey are reported in Figure 6. While the overall deformations depicted in Figure 4 are minimal, the schematics in Figure 5 qualitatively establish that regional alterations were more substantial. When the regional inter-eye difference data for each position and thickness parameter are arranged by monkey (Figure 6) three patterns of change are apparent. Monkey 1 demonstrates minimal regional position or thickness change. Monkeys 2, 3 and 4 demonstrate non-colocalized regional thinning and peripapillary scleral bowing. Monkeys 5 and 6 demonstrate colocalized regional thinning and position changes. These three patterns of early IOP-induced alterations, and the fact that each includes minimal laminar position change, are the principal findings of this report.
While the high IOP eye of monkey 1 demonstrated overall peripapillary scleral thickening (Table 2) and a region of colocalized neural canal thinning and peripapillary scleral posterior bowing and thickening (Figure 6), these changes were of minimal magnitude and extent.
Monkeys 2, 3 and 4 demonstrate EPIDmax thinning of the neural canal and lamina and non-colocalized peripapillary scleral bowing.
Monkeys 5 and 6 demonstrate enough regional position and thickness change so that these phenomena colocalize. The phenomenon of colocalization is not precise, but is best seen in the regions of EPIDmax post-NCO total prelaminar volume expansion that broadly colocalize with regional neural canal expansion and thinning as well as peripapillary scleral bowing in each eye.
There was no evidence for increased deformation in the IOP-45 (monkeys 1-3) compared to the IOP-30 (monkeys 4-6) eyes after 15 minutes of acute IOP elevation, either overall or regionally. However, the most extensive deformation clearly occurred within the one animal (monkey 6) in which the acute IOP elevation to 30 mmHg was held for 30 rather than 15 minutes.
The purpose of this study was to three-dimensionally quantify the laminar, neural canal wall and peripapillary scleral components of ONH connective tissue deformation following 15 minutes of acute IOP elevation in 5 monkeys (30 minutes in monkey 6). The principal findings of this report are as follows. First, in all animals, the lamina deformed minimally in the high IOP eye, and was not only posterior but in some cases anterior. Second, all deformations (both position and thickness changes) were regional and specific to individual eyes. Third, the regional deformation within the six high IOP eyes spanned the following spectrum: minimal detectable change in monkey 1; regional thinning accompanied by non-colocalized peripapillary scleral bowing in monkeys 2, 3 and 4; and colocalized regional thickness and position changes in monkey 5 (superior temporally and temporally) and monkey 6 (inferior temporally and temporally). Fourth, by far the greatest deformation occurred in the one eye subjected to 30 minutes of elevated IOP. Fifth, neither the magnitude of IOP elevation nor the architecture of the ONH connective tissue appeared to relate to the magnitude of deformation.
These findings should be viewed in the following context. First, the lack of substantial laminar deformation in these animals is compatible with our previous study19, that of Levy and Crapps22 in which the period of IOP elevation was short, and recent studies in living human eyes using SD-OCT by Agoumi (Y. Agoumi et al. IOVS 2009;49:ARVO E-Abstract 4898). These findings are also consistent with predictions made by computational models done by Sigal et al. 51. Using eye-specific models, Sigal and colleagues observed that an acute increase in IOP led to peripapillary sclera bowing and lateral displacements of the lamina (mostly in the periphery) while the central lamina remained stationary or displaced slightly anteriorly, depending on the eye. The study, however, did not address the differences in ONH architecture that led to the variations in response to IOP.
Our deformations are less than those reported by Yan, et al 21 in which IOP was increased for 24 hours. Laminar and scleral canal wall deformations are likely to be viscoelastic23, 52, meaning the tissue reaches full deformation only after a given period of time, so longer IOP elevations may induce larger deformations. Although ex-vivo studies of human eyes suggest that an equilibration time of 15 minutes is sufficient to obtain a stable deformation of the vitreo-retinal interface 53, it is possible that the deformations in this study may have been more substantial had the period of IOP elevation been extended to hours or days. Based on our previous studies of connective architecture in monkey eyes19, 48 and recent parameterized finite element modeling of human51, 54-56 and monkey (Sigal IA, et al. IOVS 2008;49:ARVO E-Abstract 3668, Roberts MD, et al. IOVS 2008;49:ARVO E-Abstract 3669) ONHs, we believe that the manner in which the lamina cribrosa and/or scleral canal deform following a given IOP elevation is determined by the level and duration of IOP elevation and the structural stiffness of each tissue.
The structural stiffness of a tissue is the combination of its architecture (the quantity and distribution of load-bearing tissue) and its material properties (the stiffness or compliance of the tissue), both of which contribute to a structure's ability to withstand deformation under applied load 57. The results obtained in this work suggest that the structural stiffness of the lamina cribrosa and peripapillary sclera interact (Figure 7), which is consistent with our previous report 48 and predictions made with computational models57,65. For analysis it is useful to separate the IOP-induced deformation of the ONH tissues into scleral canal expansion or contraction, laminar and scleral canal thinning, and anterior or posterior laminar and peripapillary sclera deformation. Within this lamina-sclera dynamic, the structural stiffness of the peripapillary sclera largely governs whether the scleral canal expands following an increase in IOP. When the structural stiffness of the sclera is relatively low, an increase in IOP would produce an expansion of the canal, pulling on the lamina – an indirect effect of IOP on the lamina itself. When, instead, the structural stiffness of the sclera is relatively high, an increase in IOP produces a much more modest expansion of the canal. In this case the lamina cribrosa is not pulled taut, and is left to resist the direct force of IOP with its structural stiffness alone, deforming posteriorly. We believe that in most eyes the situation would not be at either of the extremes described above, but rather some combination of the two cases. It should be noted that even in the setting of minimal anterior or posterior laminar deformation, substantial tensile strain (local deformation) within the laminar beams may be induced by the expanding scleral canal 51.
Second, the fact that all deformations (both position and thickness changes) were regional and specific to each eye is best appreciated within the regional data of Figure 6. In saying that the deformations were regional, we mean that the EPIDmax deformations were focal (only one region) or regional (at least two regions), or they were greatest within a single region when diffuse (monkey 6 only). By individual-eye specific, we mean that the regions of greatest deformation were not consistent within these six high IOP eyes. We believe our study is the first to establish that early ONH connective tissue deformation following acute IOP elevation is regional in nature and is likely to be individual-eye specific.
Third, we have observed three patterns of acute ONH connective tissue deformation within these 6 animals: 1) minimal deformation; 2) non-colocalized regional thinning and peripapillary scleral bowing; and 3) colocalized position and thickness change. To determine if the observed deformations are portions of a continuum of ONH connective tissue deformation will require the development of spectral domain optical coherence tomography49, 58, 59 (Y. Agoumi et al. IOVS 2009;49:ARVO E-Abstract 4898) or second harmonic imaging60 to capture these same ONH landmarks at multiple time points following acute IOP elevation in living human and monkey eyes. Regardless of whether these patterns are part of a continuous spectrum of connective tissue deformation, our data strongly suggest that laminar and neural canal thinning, neural canal expansion, and peripapillary sclera bowing are early responses to acute IOP elevation in the monkey eye and that, within the limits of our experimental techniques, these phenomena colocalize within those eyes demonstrating the greatest deformation.
Fourth, by far the greatest deformation occurred in the one eye subjected to 30 minutes of elevated IOP even though pressure was only raised to 30 mmHg. This animal, which was among the younger animals, but not the youngest, had the thickest lamina and peripapillary sclera and the smallest scleral canal (Table 2 and Figures 4 and and5).5). It could be that the duration of IOP exposure may be an important determinant of acute ONH connective tissue deformation. Several investigators have reported substantially increased ONH connective tissue deformations following the prolonged application of load21, 61, 62. It is possible that the additional duration of IOP elevation and time dependant (or viscoelastic) effects alone, account for the magnitude of deformation in monkey 6. However, without more eyes exposed to longer IOP elevations and/or a series of individual eyes imaged at multiple time points following acute IOP elevation, the data from this one animal do not prove such a viscoelastic effect.
It is also possible that monkey 6 may simply have had a much more compliant lamina and sclera than the other animals. In our previous histologic study of ONH connective tissue compliance19, none of the normal monkeys achieved this magnitude of deformation, though some experienced up to 80 minutes of IOP elevation. This, along with the fact that this animal appears to have the thickest lamina and sclera and smallest scleral canal (Table 2 and Figures 4 and and5),5), suggests that the material properties of both the lamina and sclera in this eye are substantially less stiff than the other 5 high IOP eyes. Estimates of these material properties within engineering finite element models63, 64 will be the subject of a future report.
Fifth, neither the magnitude of IOP elevation nor the tissue-level architecture of the ONH connective tissue closely correlates to the magnitude of connective tissue deformation. While we believe that there should be consistent biomechanical determinants of ONH connective tissue behavior, these relationships are not obvious from the histomorphometric measurements in this report. Elucidating these relationships is one of the goals of ongoing finite element modeling54,64,65.
Our findings are limited by the following considerations. While we recently improved our 3-D histomorphometric methodologies to acquire data using a 1.5 × 1.5 × 1.5 μm voxels28, all of the reconstructions for this report were acquired with our older 2.5 × 2.5 × 3.0 μm voxel technique. Most of the differences we report substantially exceed the resolution of these 3-D reconstructions and we have previously demonstrated good reproducibility of our delineation techniques at this resolution24-26. In addition, there is tissue shrinkage effects from both fixation and embedding associated with this technique, but comparisons between the two eyes of each monkey should be valid since all eyes were treated identically.
For this study, we chose acute IOP elevations of 15-minute duration for two reasons. First, our previous ONH surface compliance testing had suggested that the majority of ONH surface deformation during 60 minutes of acute IOP elevation occurred during the first 15 minutes of elevated IOP66, 67. Second, perfusion fixation is done under deep pentobarbital anesthesia, which can lower systemic blood pressure. In previous animals, we had seen evidence of poor perfusion of the central retinal vessels in eyes that were perfusion fixed after 30 to 80 minutes of high IOP. However, even in the 5 high IOP eyes of this report using our old sacrifice protocol, blood pressures were low at the time of perfusion and we saw evidence of residual blood within the retinal and posterior ciliary circulation that suggested less than ideal fixative perfusion. This residual blood may also have been due to the fact that the fixative perfusion pressure may not have been high enough (3-4 psi) to push the fixative fluid into the small arteries of the ONH due to the flow limiting nature of the old perfusion system.
While all eyes sat at their set IOP for 60 minutes after perfusion, it is possible that our data may underestimate the magnitude of acute IOP induced deformation of ONH connective tissue if fixation was less than ideal. However, monkey 1 was sacrificed using our most current protocol, which ensures a fixative perfusion pressure of 80 mmHg as measured by direct real-time pressure monitoring via brachial artery cannulation. We did not see larger connective tissue deformations in this monkey, which suggests that the small deformations we report for the other monkeys are not artifactual.
Our study may also underestimate the laminar component of ONH connective tissue deformation following acute IOP elevation if cerebrospinal fluid (CSF) pressure became elevated during the period of perfusion, and retrolaminar tissue pressure was inadvertently elevated as a result 68. We did not monitor the CSF pressure in these experiments; however, we think this phenomenon is unlikely for the following reasons. First, as outlined earlier, BP during perfusion was probably low in the last five eyes and was not greater than 80 mm Hg in the first. Although the relationship between BP change and CSF pressure change in an animal that is undergoing fixation with glutaraldehyde has not been characterized, the fact that BP was low in most eyes suggests that CSF pressure was also likely low. Second, we qualitatively compared the size of the anterior-most portion of the subarachnoid space within the 3D histomorphometric reconstructions of the other 5 monkeys (perfusion fixed at low BP) to the first monkey and found no obvious difference. This suggested that the volume of the CSF space was not altered by the higher perfusion BP in the first animal. Finally, even if CSF pressure slowly climbed during the period of perfusion fixation, this process would have occurred at the same time the connective tissues were being fixed into a position determined by the previous 15 minutes of IOP elevation. We doubt that a slow CSF rise would alter the effects of glutaraldehyde being delivered to the tissues via the blood vessels.
However, a qualitative inspection of the subarachnoid space in all of our perfusion fixed and immersion fixed tissues 19, 20, 24-26, 28 suggests that the subarachnoid space is expanded in the perfusion fixed eyes. While the collapsed (immersion fixed) vs expanded (perfusion fixed) subarachnoid space may only represent the effects of physiologic CSF pressure (in the perfusion fixed eyes), elevated CSF pressure in the animals of this study affecting the laminar (but not the scleral) deformations we report remains a possibility, and may require further investigation.
It is possible that trephinating a 6-mm-diameter piece of the ONH and peripapillary sclera allowed residual stresses within the sclera to be relieved. The resultant post-fixation tissue warping, if present, could induce non-IOP-related alterations in tissue position that would confound our measurements. However, these effects would most likely occur in the peripheral sclera closest to the trephine edge, and therefore are unlikely to affect the measurements within the ONH and peripapillary sclera. In addition, we feel that a large degree of tissue warping in the setting of months of glutaraldehyde fixation is unlikely.
All of our measurements were made relative to a plane based on the delineated NCO points (which are the end of Bruch's membrane in most monkeys)24. Any collapse of the Border Tissue of Elshnig or compression of the choroid either from the period of high IOP or from post-trephination tissue warping, could conceivably alter these points leading to a posterior shift or tilt of the reference plane in the high IOP eye, which would induce anterior movement or tilting of the other structures relative to its contralateral normal. In the case of reference plane tilt, the effect would be greatest in the periphery and least near the point of inflection. In the case of a posterior shift, the effect would be similar in both central and peripheral measurements.
We believe that the measurements we report for both eyes of each animal are accurate based on the previously reported reproducibility of our delineation method and the voxel dimensions within our reconstructions24-26. We do not believe that important shifts in the reference plane occurred as a result of focal alterations in Border Tissue of Elschnig architecture or choroidal compression within the high IOP eyes for the following reasons. First, overall thinning of the anterior neural canal opening (ASCO and ALI) achieved statistical (but not EPIDmax) significance in monkey 3 only. Regional EPIDmax thinning of the anterior neural canal was only present in monkey 3 and 6. A small tilting effect can also be seen within the regional cross section schematics for monkeys 3 and 6 (Figure 5), wherein the high IOP ONH seems slightly rotated relative to the low IOP ONH.
The lack of substantial anterior scleral canal thinning in all 6 animals is important because it suggests that little posterior deformation of NCO relative to the ONH and peripapillary scleral landmarks we studied (as might happen if the choroid was compressed, or the Border Tissue of Elschnig was deformed by the acute IOP elevations) is present in any of the 6 high IOP eyes. The fact that the regional EPIDmax posterior canal thinning detected in Monkeys 1, 2, 5 and 6 greatly exceed the anterior canal thinning within their same regions (significant or otherwise) strongly suggests that the detected posterior canal thinning in these animals are not measurement artifact.
Second, we closely examined the 0°, 30°, 60°, 90°, 120° and 150° digital section images from both eyes of each animal and found no qualitative evidence of choroidal compression in the high IOP eyes. In fact, in all monkeys the choroid is substantially expanded even in the face of the elevated IOP in the high IOP eye.
Our study includes 5 rhesus and one cynomolgous monkey which may confound our results. While there may be species differences in normal monkey ONH connective tissue architecture and material properties that could influence their response to acute IOP elevation19, 20, 69, we doubt these are important in our study for the following reasons. First, by qualitative comparison the normal eye measurements of monkey 4 (cynomolgous) fall within the range determined by the 5 rhesus animals for each parameter except for NCO offset and laminar position where they are only minimally out of that range (Table2). Second, the overall behavior in the high IOP of this animal is also within, rather than at one extreme of the five rhesus animals (Figure 4 and and66).
We did not perform a formal assessment of reproducibility on the eyes in this study and the small differences we report may be within the variability of our measurements. However, we think this is unlikely for most parameters for the following reasons. First, the intra- and inter-delineator reproducibility of our parameters (Table 2) has been extensively characterized in a series of previous reports24-26. Second, close inspection of these data suggest that the range of intra-delineator variability for most parameters is less than the PID maximum value for each parameter28 that was used as the criteria for EPIDmax change. Thus, we believe that for most parameters, our EPIDmax difference, when present, represents change beyond the variability due to delineation, had it been directly assessed in these eyes.
Finally, we choose to order our monkeys (1-6) based on the regional change in the parameter Post NCO – Total Prelaminar Volume (Table 2 and Figure 6) and the degree to which this co-localized with EPIDmax changes in other parameters. Using these criteria, monkeys 5 and 6 demonstrate the greatest amount of co-localized deformation. At present, there is no precise way to order the monkeys deformation and this choice may be arbitrary, as by other overall and regional measures other monkeys may exceed Monkey 5. Our choice of this parameter reflects our underlying belief that there is a progression of the individual forms of deformation within these tissues that leads to their co-localization. At present this is a hypothesis which needs to be studied in living eyes using in-vivo measures of ONH connective tissue deformation as mentioned above.
Our results have the following implications. First, while the lamina demonstrated little anterior or posterior deformation it did demonstrate regional thinning, in some cases colocalized, with the greatest magnitude of canal expansion (monkey 6, Figure 6). This point is important because it emphasizes the difference between overall deformation of a biological structure (the lamina cribrosa) and local deformation (strain) within its component elements. Expansion of the scleral canal and the associated laminar thinning should result in expansion of the laminar pores and thinning and elongation of laminar beams. The effects on the contained laminar capillaries, astrocytes, and adjacent axons, as well as the transfer of nutrients between them may be substantial. Thinning of the lamina cribrosa also increases the steepness of the translaminar pressure gradient ((IOP – retrolaminar tissue pressure)/laminar thickness), which may have separate implications on astrocyte physiology and retinal ganglion cell axoplasmic transport and flow15, 16, 68, 70-75. Characterization of alterations in laminar beam microarchitecture27, 63 (Grimm J, et al. Burgoyne CF, IOVS 2007;48:ARVO E-Abstract 3295) in these eyes, and the stresses and strains contained therein76 (Kodiyalam S, et al.IOVS 2005;46: ARVO E-Abstract 1267), will also be the subject of future reports.
Second, while peripapillary scleral bowing and scleral canal expansion are likely to be important determinants of lamina cribrosa biomechanical behavior, their direct and indirect effect on axonal susceptibility within the ONH remains unclear. For a given perfusion pressure, bowing and thinning of the peripapillary sclera and laminar beams should influence blood flow within the contained laminar capillaries and posterior ciliary arteries, but these effects have yet to be determined.
Finally, our EPIDmax criteria for post-mortem change detection may be conservative. Within the regional data, in many instances statistically significant differences did not achieve EPIDmax criteria and were not considered as change. We believe the trends that are present in our data are important indicators of early viscoelastic deformation. This prediction is currently under study using in vivo SD-OCT imaging of the neural canal, lamina cribrosa and peripapillary sclera after 30, 60, 90 and 120 minutes of IOP elevation in very young and very old monkey eyes (Burgoyne CF, et al. IOVS 2008;49:ARVO E-Abstract 3655).
The authors gratefully acknowledge the following individuals for their assistance with this study. Jonathon Grimm and Juan Reynaud for their assistance with software for volumetric and thickness quantification, Galen Williams and Erica Dyrud for their help with delineation, Pris Zhou and Anthony Bellezza, PhD for their direction of animal testing, Joanne Couchman for her assistance with manuscript preparation and Stuart Gardiner for his statistical consultation.
Supported in part by USPHS grants R01EY011610 (CFB) from the National Eye Institute, National Institutes of Health, Bethesda, Maryland; a grant from the American Health Assistance Foundation, Rockville, Maryland (CFB); a grant from The Whitaker Foundation, Arlington, Virginia (CFB); a Career Development Award (CFB); The Legacy Good Samaritan Foundation, Portland, Oregon; and the Sears Trust for Biomedical Research, Mexico, Missouri.