Corneal Edema Recovery Dynamics in the Rabbit

Investigative Ophthalmology & Visual Science, Vol. 31, No. 10, October 1990 Copyright Association for Research in Vision and Ophthalmology Corneal Edema Recovery Dynamics in the Rabbit A Useful Model? Perer R. Herse Open-eye and closed-eye recovery from contact lens-induced corneal edema was measured on the right eyes of 10 New Zealand White rabbits using ultrasound pachometry. Edema was produced by either 2 hours of eye closure (Patch Test) or 2 hours wearing of a thick hydrogel contact lens over the closed eye (Lens Test). The measured corneal edema recovery data were analyzed by both linear-regression analysis and nonlinear-regression analysis based on a previously published exponential open-eye edema recovery model. Linear-regression analysis revealed that the initial rabbit open-eye edema recovery rate of 36.3 ±11.9 (standard deviation) /um/hr was not significantly different (P > 0.25) from the initial young human open-eye edema recovery rate of 35.6 ± 3.4 jim/hr. Similarly, the initial rabbit closed-eye edema recovery rate of 14.3 ± 5.4 /um/hr was found to be not significantly different (P > 0.25) from the initial young human closed-eye edema recovery rate of 15.0 ± 2.2 jim/hr. Corneal edema recovery indices derived using nonlinear-regression analysis also had a strong similarity between rabbit and human corneal edema recovery rates. For example, the calculated percent recovery per hour (PRPH) for rabbit open-eye corneal edema recovery of 41.4 ± 11.3%/hr agreed well with the reported human open-eye PRPH of 34.2-58.9%/hr. The rabbit closed-eye PRPH of 25.6 ± 10.7%/hr was also found to lie between the calculated human closed-eye PRPH values of 19.9-30.2%/hr. The close similarity in both open-eye and closed-eye edema recovery values for human and rabbit cornea offers strong support for the use of the rabbit cornea as a model for human corneal edema recovery dynamics. Invest Ophthalmol Vis Sci 31:2003-2007, 1990 The rabbit cornea has been used extensively as a model in the study of corneal physiology. 1 " 4 In particular, the rabbit has been commonly used for in vitro studies of corneal hydration control. 5 " 8 The validity of the rabbit model has been questioned, however, after a report that the rabbit corneal endothelium, unlike that in humans, has a remarkable capacity to regenerate. 9 A counterargument can be found in a report stating that the regenerative capacity of the rabbit endothelium is artifactual due to the use of immature animals. 10 " Thus, there is some uncertainty as to the validity of using the rabbit as a model for corneal hydration control studies. Numerous reports propose that corneal hydration control can be assessed in vivo by measurement of corneal recovery from a particular level of stress-induced edema towards normal thickness. 12 " 18 Two analysis methods were developed from these studies: (1) linear-regression analysis 16 " 18 and (2) nonlinearregression analysis. 1314 Li near-regression analysis estimates the initial rate of corneal edema recovery during the first stages of edema recovery. The useful- Reprint requests: Peter R. Herse, Dipp App Sc (Optom), PhD, Department of Optometry, University of Auckland, New Zealand. ness of the linear-regression method is limited because it is unable to explain the nonlinear edema recovery curve seen experimentally. Nonlinear-regression analysis, on the other hand, offers a more realistic analysis model because it assumes that corneal deswelling follows an exponential path, whose shape may be determined by a small number of mathematic constants. This model is more clinically applicable; it can provide estimates of a number of corneal hydration parameters such as the corneal edema recovery rate, baseline corneal thickness, and the time required to recover to a particular fraction of the level of initial edema. Both of these models have been used to estimate corneal hydration control parameters in normal, aged, and diseased in vivo human corneas. 12 " 1518 No studies have as yet been done to assess the edema recovery characteristics of the in vivo rabbit cornea to our knowledge. We tried to quantify the open-eye and closed-eye edema recoveries of the in vivo rabbit cornea using both analysis methods. This study compared rabbit data with previously published human corneal edema recovery parameters and assessed the validity of the rabbit cornea as a model for recovery of the human corneal from induced edema. 2003

2004 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Ocrober 1990 Vol. 31 Subjects Materials and Methods Ten 19-week-old male New Zealand White rabbits were obtained from a local supplier (Ray Nichols, Lumberton, TX). The rabbits were housed in the Animal Care Unit of the University of Houston (AALAC approved) under a 12-hr light/12-hr dark cycle (lights on at 7 AM). Animals were allowed free access to food (Purina Rabbit Chow, Ralston-Purina, St. Louis, MO) and water, and were handled in accordance with the ARVO Resolution on the Use of Animals in Research. Corneal Thickness Measurement Central corneal thickness measurements were made by lightly touching the water-filled probe of an ultrasound pachometer (Storz Corneo-Scan 2000, St. Louis, MO) to the tear film of the rabbit cornea and obtaining six readings whose mean was used as the recorded corneal thickness measurement. No anesthetic was used, since the rabbits showed no discomfort during this procedure. The probe was hand-held and the position of the central cornea judged visually. This procedure has been previously shown to be valid, 19 " 21 particularly so in the rabbit due to the relative constancy of the rabbit central corneal thickness. 19 The ultrasound tissue velocity was set at 1580 m/sec. 19 Baseline Corneal Thickness Because these laboratory rabbits tended to have irregular sleep patterns, baseline corneal thickness was determined by measuring the right eye corneal thickness of each rabbit every 2 hr for 24 hr. The average of these readings was used as an estimate of the baseline corneal thickness. Edema Production Patch test: Corneal edema was induced by patching both eyes closed for 2 hr. At the end of this time, the right eye was briefly opened, and the corneal thickness quickly measured. This procedure was done on all ten rabbits. Lens test: Corneal edema was produced by inserting a thick soft hydrogel contact lens (+11.00B3; Bausch & Lomb, Rochester, NY) over the closed right eye for 2 hr. At the end of this time, the eye was briefly opened, the lens removed, and the corneal thickness quickly measured. This procedure was also done on all ten rabbits. Edema Recovery Open-eye: Open-eye edema recoveries were measured after both the Patch Test and Lens Test methods of producing edema. After inducing corneal edema, the eye was opened and the corneal thickness measured each 30 min for 3 hr (ie, seven measurements taken during the deswelling period). This procedure was done on all ten rabbits. Closed-eye: Closed-eye edema recoveries were measured after both the Patch Test and Lens Test methods of producing edema. After inducing corneal edema, the eye remained closed and was opened briefly only during the measurement process. Corneal thickness measurements were taken hourly for 4 hr (ie, five measurements taken during the deswelling period). This procedure was also done on all ten rabbits. Analysis Method Linear-regression analysis: In the linear-regression analysis method, 16 " 18 it is assumed that the initial phase of edema recovery can be modeled by a linear decrease of corneal thickness over time (/xm/hr). The analysis was done during the first 1.5 hr of open-eye edema recovery and during the first 2 hr of closed-eye edema recovery. In both instances the correlation coefficient of the linear regression was found to be greater than 0.90. Nonlinear-regression analysis: A mathematic model has been proposed for human open-eye corneal edema recovery. 1314 The model can be expressed as: T = B + S,exp- Dl) + e where T is the instantaneous corneal thickness (^m), B is the baseline corneal thickness (/im), S, is the induced corneal swelling (jum), D is a calculated deswelling factor, t is the time after the start of the edema recovery (min), and e represents the error term of the mathematic model (ie, measurement error). This expression was shown to predict the open-eye edema recovery adequately in normal human subjects of various ages and in elderly subjects with Fuch's corneal endothelial dystrophy. The expression was also shown to provide clinical indices of corneal hydration control such as: (1) a corneal deswelling rate in percent recovery per hour (PRPH), (2) a time to remove 95% of the initial edema (T 95% ), and (3) the open-eye steady state corneal thickness (OESS). These edema recovery constants can be calculated by entering the raw corneal thickness data from the edema recovery experiments into a nonlinear-regression procedure (Marquardt method; SAS NLIN, Cary, NC). For a more detailed explanation, see Mandell et al 13 and Poise et al. 14 Statistical analysis: Analysis of variance was used to test statistical differences between group means.

No. 10 RABBIT CORNEAL EDEMA RECOVERY DYNAMICS / Herse 2005 Results Baseline Corneal Thickness The mean rabbit corneal thickness over the 24-hr period was found to be 350 ± 10 /urn. This agrees well with literature values. 22 ' 23 All values reported here represent the mean of the experimental measurements ± one standard deviation. 440-6 PATCH TEST O LENS TEST Edema Production Combining the data obtained from the open-eye and closed-eye procedures, it was found that the 2-hr insertion of the thick hydrogel contact lens (Lens Test) increased rabbit corneal thickness to 428 ± 16 /urn, producing 22% edema. Similarly, 2 hr of eyeclosure (Patch Test) in the rabbit increased corneal thickness to 391 ± 16 //m, producing 12% edema. The edema induced by the Lens Test was significantly greater (P < 0.0001) than that induced by the Patch Test. Edema Recovery Open-eye edema recovery: The open-eye edema recovery functions from both the Lens Test and Patch Test are shown in Figure 1. The corneal thickness is seen to deswell in a nonlinear manner in both conditions until an asymptotic baseline is reached after approximately 3 hr at approximately 356 ± 11 ^m. Closed-eye edema recovery: The closed-eye edema recovery functions from both the Lens Test and Patch Test are shown in Figure 2. The closed-eye recovery from the Lens Test underwent a nearly linear decrease in corneal thickness over 4 hr after the removal of the edema stimulus. In the Patch Test, the eye remained closed for 6 hr. The cornea was seen to reach maximal thickness 3-4 hr after eye closure; o LENS TEST PATCH TEST Fig. 2. Normal rabbit closed-eye edema recovery (n = 10). Error bars represent ±1 SEM. Measurements taken at 60-min intervals for 4 hr. thereafter a relatively constant corneal thickness was attained. Linear-regression analysis: The rate of open-eye edema recovery during the first hour after the Lens Test was calculated to be 36.3 ± 11.9 /um/hr. The closed-eye edema recovery rate during the initial 2 hr of edema recovery after the Lens Test was calculated to be 14.3 ± 5.4 jim/hr. These data are shown in Table 1. Nonlinear-regression analysis: The test data were analyzed using the exponential open-eye edema recovery model. 1314 The calculated deswelling constants are listed in Table 2. After rejecting steady-state corneal thickness estimates with asymptotic standard errors greater than 5 jim, the rabbit OESS corneal thickness was calculated to be 353 ± 14 ^m, in good agreement with the baseline estimate of corneal thickness of 350 ± 10 /urn. Due to the greater variability noted in the closed-eye edema recovery data, closed-eye steady state (CESS) estimates having asymtotic standard errors greater than 10 ixm were rejected. These rejection criteria are more stringent than the 15-/um rejection criteria used by Poise et al. 14 When this was done, the rabbit CESS corneal thick- Table 1. Corneal edema recovery rates in /um/hr (mean ± 1 SEM) derived using linear regression analysis using normal human data from the literature and experimental rabbit data (n = 10) Eve Edema recovery (u Fig. 1. Normal rabbit open-eye edema recovery (n = 10). Error bars represent ±1 SEM. Measurements taken at 30-min intervals for 3 hr. Open Rabbit Closed Rabbit 26.5 ±2.2 36.3 ± 3.8 35.6 ± 1.1 10.5 ±0.9 14.3 ± 1.7 15.0 ±0.7

2006 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Ocrober 1990 Vol. 01 Table 2. Corneal edema recovery constants (mean ± 1 SEM) derived using the exponential deswelling model using normal human data from the literature and experimental rabbit data (n = 10) Eye Open Old human 14 Normal rabbit Young human 14 Closed Normal rabbit PRPH (%/hr) 34.2 ± 2.3 41.4 ±3.6 58.9 ±2.8 19.9 25.6 ± 3.4 30.2 T 9S % (hr) 7.5 ±0.6 6.2 ± 0.8 3.5 ±0.3 13.5 12.6 ±2.3 8.3 SS(ti) m) 537 ± 10 353 ± 4 516± 12 388 ± ness was calculated to be 388 ± 9 ^m, in good agreement with the closed-eye corneal thickness of 391 ± 22 nm found after 6 hr of eye closure in the Patch Test. The edema recovery indices (PRPH and T 95% ) were determined mathematically using the following expressions: 1314 PRPH= 100(1 -exp" 60D ) T 95% = 2.996/D The rabbit open-eye corneal PRPH and T 95% were calculated to be 41.4 ± 11.3%/hr and 6.2 ± 2.5 hr. The rabbit closed-eye corneal PRPH and T 95% were calculated to be 25.6 ± 10.7%/hr and 12.6 ± 7.2 hr. Discussion A major factor in the use of an animal model is its applicability to the human condition. Figure 3 illustrates the open-eye edema recoveries for rabbit and human 1417 corneas. The close similarity of the rabbit and human open-eye edema recovery functions support the proposal that the rabbit may provide a useful model for corneal deswelling dynamics. Further support may be obtained by considering the results of the mathematic analysis of the edema recovery data. Using linear-regression analysis, for the first hour, the young human (mean age, 27 yr) 17 and rabbit open-eye edema recoveries were found to be similar (P > 0.25) at 35.6 ± 3.4 Mm/hr and 36.3 ±11.9 Aim/hr. The closed-eye edema recoveries for young human 17 and rabbit corneas, for the initial 2 hr of deswelling, were also similar (P > 0.25) at 15.0 ± 2.2 /um/hr and 14.3 ± 5.4 ^im/hr. Further data supporting the similarity between human and rabbit corneal deswelling rates can be found by comparing the linear-regression open-eye and closed-eye deswelling rates found for the rabbit with human data from varying age groups. These data are shown in Table 1. The rabbit open-eye deswelling rate fell near the value reported for the young 4 (mean, 27 yr) human and the old (mean, 66 yr) human. 17 Similarly, the rabbit closed-eye deswelling rate was found to lie between the young human and old human deswelling rates. These similarities, as found by linear-regression analysis, further strengthen the use of the rabbit cornea in studies of corneal edema recovery dynamics. The human and rabbit corneal edema recovery constants calculated using nonlinear-regression analysis are listed in Table 2. It can be seen that the openeye and closed-eye edema recovery constants for the rabbit cornea agree closely with human values, with the rabbit constants lying between those reported for old (mean, 72 yr) and young (mean, 24 yr) human subjects. 14 This result, when combined with the previous results, offers strong support for the use of the rabbit as a model for studies of corneal edema recovery. On a clinical level, our investigation supports the claims of Mandell et al 13 and Poise et al 14 that the exponential edema recovery model may enable clinical assessment of in vivo corneal hydration control. With the current availability of fast and reliable ultrasound pachometers, we can foresee immediate clinical applications in contact lens wear and pathology detection. Unfortunately, and in agreement with Mandell et al 13 and Poise et al, 14 a great deal of refinement is required before this test can become a clinically useful tool. At present, the long testing durations and computer-intensive analysis techniques restrict this method to a laboratory setting. In conclusion, we provided evidence to support the use of the human open-eye exponential edema recovery model 1314 for the analysis of human closed-eye corneal edema recovery. The model has also been shown to be useful in describing both the open-eye RABBIT HUMAN (O'Neal & Poise, 1986) HUMAN (Poise et al, 1989) Fig. 3. Open-eye edema recovery functions for human and rabbit cornea. The curve represents the best-fitting third-order polynomial through the rabbit data. The error bars are ±l SEM. The human data are from O'Neal 17 and Poise et al. 14 Measurements taken at 30-min intervals for 3 hr.

No. 10 RABBIT CORNEAL EDEMA RECOVERY DYNAMICS / Herse 2007 and closed-eye edema recovery in rabbit cornea. Although an extremely powerful laboratory tool for assessment of in vivo corneal hydration control, a great deal of refinement of the analysis techniques and experimental design is necessary before this method may be used clinically. The close similarities between human and rabbit corneal edema recoveries, using both linear- and nonlinear-regression analysis, strongly suggest that the rabbit cornea offers a useful model for the study of corneal edema recovery dynamics. Key words: cornea, deswelling, pachometry, rabbit Acknowledgments The author thanks Drs. Donald Pitts and Roger Boltz for their editorial assistance, John Feaster for his help in running the SAS:NLIN procedure, and the University of Houston Institute for Contact Lens Research for providing the financial support to enable completion of this project. References 1. Burstein NL and Klyce SD: Electrophysiologic and morphologic effects of ophthalmic preparations on rabbit corneal epithelium. Invest Ophthalmol 16:899, 1977. 2. Chang EP, Kedy DA, and Chien JCW: Ultrastructure of rabbit corneal stroma: Mapping of optical and morphological anisotropes. Biochim Biophys Acta 343:615, 1974. 3. Dahl IMS and Coster L: Proteoglycan biosynthesis in cultures of corneas and corneal stroma cells from adult rabbits. Exp Eye Res 27:175, 1978. 4. Farrell RA, McCally RL, and Tatham PER: Wavelength dependencies of light scattering in normal and cold swollen rabbit corneas and their structural implications. J Physiol 233:589, 1973. 5. Mishima S and Kudo T: In vitro incubation of rabbit cornea. Invest Ophthalmol 6:329, 1967. 6. Maurice DM: The location of the fluid pump in the cornea. J Physiol 221:43, 1972. 7. Fischbarg J, Lim JJ, and Bourguet J: Adenosine stimulation of fluid transport across the rabbit corneal endothelium. J Membr Biol 35:95, 1977. 8. Anderson El, Fischbarg J, and Spector A: Fluid transport, ATP level and ATPase activities in isolated rabbit corneal endothelium. Biochim Biophys Acta 307:557, 1973. 9. Van Horn DL, Sendele DD, Seideman S, and Buco PJ: Regenerative capacity of the corneal endothelium in rabbit and cat. Invest Ophthalmol 16:597, 1977. 10. Bito LZ: Species differences in the responses to the eye to irritation and trauma: A hypothesis of divergence in ocular defense mechanisms, and the choice of experimental animals for eye research. Exp Eye Res 39:807, 1984. 11. Baroody RA, Bito LZ, DeRousseau CJ, and Kaufman PL: Ocular development and aging: 1. Corneal endothelial changes in cats and in free-ranging and caged rhesus monkeys. Exp Eye Res 45:607, 1987. 12. Vannas A, Holden BA, Sweeney DF, Efron N, and Poise KA: Deswelling of the graft cornea following hypoxic edema. Acta Ophthalmol 62:879, 1984. 13. Mandell RB, Poise KA, Brand RJ, Vastine D, Demartini D, and Flom R: Corneal hydration control in Fuch's dystrophy. Invest Ophthalmol Vis Sci 30:845, 1989. 14. Poise KA, Brand R, Mandell R, Vastine D, Demartini D, and Flom R: Age differences in corneal hydration control. Invest Ophthalmol Vis Sci 30:392, 1989. 15. Poise KA, Mandell R, Vastine D, Demartini D, Brand R, Bonanno J, and Flom R: Corneal hydration dynamics in Fuch's dystrophy. ARVO Abstracts. Invest Ophthalmol Vis Sci28(Suppl):171, 1987. 16. O'Neal MR and Poise KA: In vivo assessment of mechanisms controlling corneal hydration. Invest Ophthalmol Vis Sci 26:849, 1985. 17. O'Neal MR: In vivo assessment of mechanisms controlling corneal hydration. AAMRL-TR-86-004, Harry G. Armstrong Aerospace Medical Center Research Laboratory, Wright-Patterson Air Force Base, OH 1986, p. 57. 18. O'Neal MR and Poise KA: Decreased endothelial pump function with aging. Invest Ophthalmol Vis Sci 27:457, 1986. 19. Chan-Ling T, Payor C, and Holden BA: Corneal thickness profiles in rabbits using ultrasonic pachometer. Invest Ophthalmol Vis Sci 24:1408, 1983. 20. Chan-Ling T, Effron N, and Holden BA: Diurnal variation of corneal thickness in the cat. Invest Ophthalmol Vis Sci 26:102, 1985. 21. Madigan MC, Gillard-Crewther S, Kiely PM, Crewther DP, Brennan NA, Effron N, and Holden BA: Corneal thickness changes following sleep and overnight contact lens wear in the primate (Macaca fascicularis). Curr Eye Res 6:809, 1987. 22. Prince JH: The Rabbit in Eye Research. Springfield, IL, Charles C. Thomas, 1964. 23. Walkenbach RJ and Corwin JG: Corneal deturgescence during organ culture. Curr Eye Res 6:381, 1987.