| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1From the Retina and Optic Nerve Research Laboratory and 2Departments of Ophthalmology and Visual Sciences, and 3Physiology and Biophysics, Dalhousie University, Halifax, Nova Scotia, Canada; and 4Alcon Research Ltd., Fort Worth, Texas.
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
METHODS. Optic nerve cross-sections from both eyes of 39 Brown Norway rats unilaterally treated with various concentrations of ET-1 or physiological saline solution via a surgically implanted osmotic minipump were processed for light and transmission electron microscopy (TEM). The optic nerve damage grade, between 0 (no damage) and 10 (total damage), was based on the number of zones of approximately equal damage and the mean percentage of damage within each zone. Grading was performed under light microscopy by three observers and compared with axonal survival determined with TEM using two quantification methods: the sampling method, in which 
10% of the section was counted, and the full-count method, in which the whole section was counted (n = 12). Axonal survival was expressed as a ratio of axon counts in the experimental to control eye. Before these comparisons, the inter- and intraobserver agreement rates were determined in another group of 85 and 12 ET-1-treated animals, respectively.
RESULTS. The interobserver 
was 0.66 (95% confidence interval [CI]: 0.58–0.74) for all eyes and 0.55 (95% CI: 0.43–0.67) for the experimental eyes only. The intraobserver was 0.80, 0.81, and 0.80 for all 24 eyes and 0.60, 0.64, and 0.71 for experimental eyes only. The correlation between damage grade in the experimental eye and axonal survival using the TEM sampling method (Spearman’s 
= –0.677 for all animals and –0.827 for the subset of animals with full counts only) was lower than that with the full-count method (Spearman’s = –0.926). When axonal survival was less than 0.7, the sampling method always underestimated the extent of damage.
CONCLUSIONS. The grading scheme had good inter- and intraobserver agreement, and high correlation with the TEM methods. It is a practical and time-saving method, requiring less than 1 minute per nerve and is an alternative to sampling methods that can yield significant errors.
Retrograde labeling of the RGCs with a tracer dye is a commonly used technique in rodents for quantifying RGCs in wholemounted retinas.10 11 For estimates based on manual counts, portions of the retina are imaged and analyzed.12 Alternatively, automated image analysis methods have been used to sample the whole retina.13 After comparison with the fellow or other control eyes, RGC survival can be estimated. Although this technique is valuable, it is relatively time-consuming if counts are performed manually. Furthermore, in long-term experiments, RGC labeling becomes faint, and some tracer dyes (e.g., Fluorogold; Fluorochrome, Denver, CO), are transported out of the RGC bodies.14 To avoid this complication, tracer dyes are often introduced after the experimental insult and just before the animal is killed,3 15 with the assumptions that labeled RGCs are viable and that all viable RGCs are labeled.
Quantifying the number of surviving RGC axons is another commonly used technique in experimental glaucoma, though the temporal relationship between axonal and somal loss may be complex (Archibald ML, et al. IOVS 2001;42:ARVO Abstract 4440).16 Cross sections of optic nerve are imaged by light or transmission electron microscopy (TEM). Similar to the quantification of RGCs, axon counts can be obtained by manual17 or automated techniques.18 19 20 Manual counting of axons is also time consuming, especially in nerves of nonhuman primates where a relatively large number of axons have to be counted compared with rodents, for a given proportion of the nerve. For these reasons, some investigators have described a qualitative or semiquantitative grade whereby optic nerve sections can be evaluated rapidly.21 22 Morrison et al.23 showed good agreement between a subjective nerve injury grade and axon counts in eight animals in whom one eye had experimental elevation of intraocular pressure (IOP).
In this study we describe a semiquantitative ordinal 11-point grading scheme using light microscopy for the evaluation of optic nerves of rats treated chronically with endothelin (ET)-1 to induce RGC loss.24 We compared this grading scheme to axon counts based on two manual methods using TEM: the sampling method, in which counts were derived from sampling approximately 10% of the nerve, and the full-count method, in which counts were derived from the entire nerve section.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Model of ET-1-Induced Chronic Optic Neuropathy
Optic nerve damage was induced using a model, detailed elsewhere,24 of RGC loss produced by a chronic delivery of ET-1. Briefly, an osmotic minipump (Durect Corp., Cupertino, CA) containing ET-1 dissolved in 0.1 mM physiological saline solution (PSS) was used to deliver 0.05, 0.1, 0.2, or 0.4 µg ET-1/d (corresponding to 3.3, 6.7, 13.4, and 26.8 µM, respectively) or PSS at a rate of 0.25 µL/hr to one optic nerve. One end of a silastic delivery tube was connected to the minipump, while the other was channeled through a hole in the orbital ridge and positioned approximately 1 mm behind the globe and just above (ensuring no contact with) the optic nerve. Each minipump contained sufficient perfusate for 28 days and minipumps were replaced where experiments exceeded this time period. The experimental procedure was performed on one eye only, and the fellow eye served as the untreated control.
Tissue Preparation and Imaging
Animals were killed at various time points after minipump implantation, as a wide range of nerve damage was desired. Previous results have indicated a time-dependent loss of RGCs and their axons.24 After enucleation, the optic nerve was cut approximately 1 mm behind the globe and the nerve stump fixed immediately in 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.2) overnight. The stump was rinsed with the buffer and placed in 1% OsO4 for 2 hours and in 0.25% uranyl acetate for an additional 2 hours. The nerves were dehydrated with a series of acetones and embedded in epoxy medium (Epon Araldite; Mirivac, Halifax, NS, Canada).
For analysis with TEM, nerves were thin sectioned (100–130 nm), poststained with 2% uranyl acetate, and viewed on an electron microscope (EM300; Philips, Eindhoven, The Netherlands). For the TEM sampling method, three or four (depending on the diameter of the nerve) equidistant micrographs (83 x 58 µm) from the center of the nerve were taken in each quadrant. For the TEM full-count method, a montage of the whole nerve section consisting of 40 to 50 micrographs was made in a subset of animals. For optic nerve grading with light microscopy, nerves were thick sectioned (1 µm) and stained with 1% p-phenylenediamine (Sigma-Aldrich, Oakville, ON, Canada). They were then rinsed in isopropanol, dried on a hotplate and coverslipped with a mounting medium (Eukitt; Electron Microscopy Sciences, Washington, PA), as previously described.25
Axon Quantification and Optic Nerve Damage Grade
Each micrograph obtained with TEM was printed on 12.5 x 12.5-cm paper with a print magnification of 4800. Axon counts were performed manually by a single observer. Criteria for viable axons included an intact myelin sheath, visible neurofilament, and absence of obvious swelling or shrinkage. The mesh grid used for TEM occupied 55% of the imaged area, and axon counts were extrapolated accordingly. Axonal survival was expressed as a ratio of axon counts in the experimental to fellow control eye.
Optic nerve grading was performed under light microscopy. Similar to the criteria used with TEM, axonal damage was identified by degeneration, swelling, shrinkage, and replacement of axons by glial tissue. Using a 10x (NA 0.3) objective (Nikon, Mississauga, ON, Canada), so that the whole nerve was visible in the field of view, we first identified the number of zones where the damage was approximately equal and then made an approximation of the percentage of the nerve cross section occupied by each zone. Using a 40x (NA 0.75) objective, under which it was possible to resolve individual axons, we approximated the mean percentage of damage (using criteria stated above) within each zone. By summing the products of the mean percentage of damage within each zone and the area of the nerve occupied by the zone, we obtained a number between 0 and 1. This number was rounded to one decimal place and multiplied by 10 to obtain the damage grade. If the calculated damage grade was 0, but the nerve contained damaged axons (typically >20), the grade was changed to 1, as a nominal indication that the nerve was damaged. Figure 1 shows illustrative examples of the grading scheme. The time required for grading a nerve was less than 1 minute and considerably less when the nerve was apparently normal.
|
Only optic nerve damage grades from the experimental nerves were used when comparisons between the grading scheme and axon counts were made. This approach was taken to avoid the artificially high correlation obtained in close data clustering when control nerves are included. To help control for the potential influence of interanimal variability in axon counts, for both TEM methods, the proportion of surviving axons in experimental to control nerves was used in comparisons with optic nerve damage grade.
Inter- and Intraobserver Variability of the Grading Scheme
Before grading the optic nerves, as described herein for the comparison study, the inter- and intraobserver variability of the grading scheme was determined for the three observers in a separate group of animals from a previous study.24 The interobserver variability was determined in a total of 85 animals in whom one eye was treated chronically with ET-1, whereas the fellow eye served as the untreated control. The nerves were divided into eight sets containing 8 to 26 nerves, with each observer grading one set per session. The intraobserver variability was determined in one set of nerves from 12 animals graded by the same observer on three different days. The nerves were coded such that the observer was unaware of the status of the nerve (experimental or control). Each observer was also masked to the grades of the other observers as well as to his or her grades from previous sessions. statistics were determined for both inter- and intraobserver variability by using the subjective classification described by Landis and Koch26 wherein 0.81 to 1.00, 0.61 to 0.80, 0.41 to 0.60, 0.21 to 0.40, 0.20 to 0.39, and less than 0 indicated almost perfect, substantial, moderate, fair, and poor agreement, respectively.26
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
The agreement rates among the three observers for the eight sets of nerves is shown in Table 1 . The mean ± SD interobserver for the eight sets of nerves was 0.67 ± 0.09. There was no obvious trend in interobserver variability with grading session (Spearman’s = 0.571; P = 0.139). The interobserver for all 170 nerves over the eight sets was 0.66 (95% CI: 0.58–0.74). When the control eyes were removed from the analyses, the agreement rates were lower (Table 1) . The mean ± SD interobserver for the eight sets of nerves was 0.55 ± 0.21, whereas the interobserver for all experimental nerves over the eight sets was 0.55 (95% CI: 0.43–0.67). The intraobserver for the three observers were 0.80, 0.81, and 0.80 for all eyes and 0.60, 0.64 and 0.71, respectively, for experimental eyes only. The interobserver agreement results are shown in Table 2 .
|
|
= –0.677; P < 0.001; Fig. 2 ); however, there was considerable dispersion in the damage grades with mild axonal loss, or there was considerable dispersion in axonal loss estimated with the sampling method for a given damage grade. The relationship between damage grade and axonal survival with the TEM full-count method is shown in Figure 3 (Spearman’s = –0.926; P < 0.001). The association between damage grade and axonal survival using the full full-count method was significantly higher than that with the sampling method (P = 0.032). When only the subset of 12 animals in whom axonal survival was obtained with both methods was analyzed, the association between damage grade and axonal survival with the full-count method (Spearman’s = –0.926) was higher than that with the sampling method (Spearman’s = –0.827, P < 0.001), although the differences did not reach statistical significance (P = 0.338).
|
|
|
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Although methods that sample the entire optic nerve cross- section ultimately provide the most accurate measures, manual counting is time consuming, even in rodents, which have less than 10% of the axons in primates. In this study, performing manual counts using the TEM sampling method required approximately 1 hour per animal, whereas the full-count method required approximately 5 hours. In experiments in which dozens of animals may be needed, manual counting is obviously impractical and tedious. Optic nerve damage grading schemes,21 22 such as the semiquantitative one described in this study may be a practical alternative.
When both nerves from a large number of animals were graded by three independent observers, the interobserver agreement was substantial. The agreement rate decreased to moderate when only experimental eyes were graded. This finding is not surprising, because the inclusion of normal fellow control eyes, which almost invariably had damage grades of 0, leads to a higher agreement rate. There is a paucity of published reports on inter- and intraobserver grading schemes for optic nerve damage. Using a 5-point (1–5) damage scale, Jia et al.21 reported of 0.90 (95% CI: 0.76–1.03), indicating almost perfect agreement, in a group of 17 animals that had unilateral experimental elevation of IOP. Although the respective in the present study was lower (0.66; 95% CI: 0.58–0.74), it is important to note that the values between the two studies may be difficult to compare, because the grading scheme described here was an 11-point (0–10) scale and that the distribution of optic nerve damage was different between the studies. Twenty-nine (85%) of the nerves in the study of Jia et al.21 had a grading of either 1 or 5, which would yield a higher . In the present study the intraobserver agreement was considerably higher than the respective interobserver agreement for all eyes and experimental eyes only. To the best of our knowledge, there are no other published reports on intraobserver agreement rates for axon-based optic nerve damage grading schemes.
Although there was a statistically significant relationship between the damage grade and axonal survival with the TEM sampling method, there was considerable variation in damage grades for a given ratio of axonal survival or vice versa. Hence, either the grading scheme does not accurately reflect the degree of axonal loss, or the TEM sampling method yields an inaccurate estimate of axonal survival. To address this question, the damage grade was correlated to the TEM full-count method. The fact that this analysis showed a considerably higher correlation between damage grade and axonal survival suggests that the TEM sampling method may yield suboptimal results. Indeed, if the TEM full-count method is assumed to be the gold standard, the sampling method yielded errors (underestimation of damage) in axonal survival that approached 0.2 in moderately damaged nerves. The grid mesh in TEM sections occupied 55% of the section; hence, even for the full-count method, only 45% of the nerve was used. As the sampling method sampled 10% of the axons, only 4.5% of the nerve was used in the counts. In normal nerves, or in nerves with completely diffuse axonal loss, sampling a small proportion of the nerve may yield reasonably accurate estimates with acceptable standard errors, as was shown recently by Cull et al.28 However, in eyes with nonuniform axonal loss across the nerve section, the undersampling is likely to lead to significant errors. In this situation, axon counts in a larger proportion of the nerve may have to be performed to increase accuracy, making the task even more time consuming.
Many improvements to the optic nerve grading scheme could have been made. For example, the addition of a 100x oil-immersion objective would have allowed resolution of finer detail and application of criteria even closer to those used for counting axons with TEM; however, we thought that the additional time and decision rules would cease to make the technique rapid. We also elected not to image and print a micrograph, as including the whole nerve section in the image would entail digitally enlarging the view with the 10x objective. Obtaining the same magnification optically produced far superior results, but capturing these images, printing micrographs, and constructing a montage for each nerve would again add significant time. For these reasons, the grading was performed under the microscope by a step-wise process with two microscope objectives. The major limitation of the grading scheme is the possibility that light microscopy may not be able to resolve the smallest axons, explaining why in previous studies total axon counts with light microscopy are lower than those obtained with TEM.23 Light microscopy may also not allow a detailed view of the axoplasm, and therefore damaged or degenerating axons that have not yet been phagocytosed may be included as viable axons in the evaluation. Although the grading scheme is quick, the ability to standardize the technique across different laboratories is unknown. Hence, for comparison of results between laboratories using grading schemes, it may be necessary to calibrate the damage grades with axonal loss using counting methods so that the results can be interpreted correctly.
In summary, we have described and evaluated a semiquantitative scheme for grading axonal damage in rat optic nerves. With modifications, this scheme can be applied to other species. Optic nerves can be graded under light microscopy in less than 1 minute per nerve, yielding high correlation to axonal survival measured with the TEM full-count method, which requires 2.5 hours. Given that axon counting with TEM is time consuming and may result in significant errors when sampling methods are used, the grading scheme may provide considerable advantages in evaluating the experimental optic neuropathies.
|
Footnotes |
|---|
Submitted for publication September 11, 2005; revised October 26, 2005; accepted December 16, 2005.
Disclosure: B.C. Chauhan, None; T.L. LeVatte, None; K.L. Garnier, None; F. Tremblay, None; I.-H. Pang, Alcon Research Ltd. (E); A.F. Clark, Alcon Research Ltd. (E); M.L. Archibald, None
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked "advertisement" in accordance with 18 U.S.C. 
1734 solely to indicate this fact.
Corresponding author: Balwantray C. Chauhan, Department of Ophthalmology and Visual Sciences, Dalhousie University, 2nd Floor Centennial Building, Queen Elizabeth II Health Sciences Centre, Halifax, NS, Canada B3H 2Y9; bal{at}dal.ca.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. D. Bull, G. A. Limb, and K. R. Martin Human Muller Stem Cell (MIO-M1) Transplantation in a Rat Model of Glaucoma: Survival, Differentiation, and Integration Invest. Ophthalmol. Vis. Sci., August 1, 2008; 49(8): 3449 - 3456. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
