Noninvasive Measurement of Rat Intraocular Pressure With the Tono-Pen

Transcription

1 Noninvasive Measurement of Rat Intraocular Pressure With the Tono-Pen C. G. Moore, Scott T. Milne, and John C. Morrison Purpose. The purpose of this study was to evaluate the Tono-Pen 2 tonometer for measuring intraocular pressure (IP) in the living rat eye. Methods. ne eye from each of 20 adult, anesthetized brown Norway rats (group 1) was cannulated and simultaneously connected to a syringe and a pressure transducer with a chart recorder. We increased IP from 15 to 45 mmhg in 5-mmHg increments and obtained 15 consecutive readings (ignoring instrument-generated averages) at each pressure increment with a Tono-Pen 2 tonometer. To test the tonopen's ability to measure unknown IP, transducer pressures were varied randomly in 20 additional animals (group 2), and tonopen readings were obtained in masked fashion. Results. Plotting the mean tonopen readings for each animal against transducer IP produced a regression formula of y = x (r = 0.98). Mean group 2 tonopen values plotted against transducer IP yielded a regression formula of y = x (r = 0.94). A method comparison analysis showed that the tonopen significantly overestimates pressures at low IP (< 15 mmhg), and it significantly underestimates pressures at high IP (> 30 mmhg). Using two-way analysis of variance, it was determined that the group 2 data did not differ significantly from the group 1 data (P> 0.76). Because of this consistency, we generated a correction factor with 95% prediction intervals for Tono-Pen readings. Conclusions. The Tono-Pen 2 can be used reliably to measure IP in the normal rat eye. Invest phthalmol Vis Sci. 1993;34: vjrlaucoma is a major cause of blindness that affects an estimated 1.6 million people in the United States. 1 In this insidious disease, optic nerve fibers are damaged at the level of the optic nerve head, usually by From the Casey Eye Institute and Portland Veterans Administration Medical Center, Portland, regon. Presented, in part, at the Annual Meeting of the Association for Research in Vision and. phthalmology, Sarasota, Florida, May Supported, by a grant from the Medical Research Foundation, of Portland, regon, by Veterans Adini.nistrat.ion grant # , and. by National Institutes of Health grant R0I EY John Morrison is an RPB Miriam and Benedict Wolfe Scholar. Submitted for publication: July 20, 1992; accepted September 21, Proprietaiy interest, category: N. Reprint requests: John C. Morrison, Casey Eye Institute, regon Health Sciences University, 3375 SW TenuiHiger Boulevard, Portland, R high intraocular pressure (IP). 23 The factors that influence the susceptibility of the optic nerve head to elevated IP are poorly understood, however. Although much has been learned from currently available acute animal models, 4 " 8 glaucoma is a chronic disease: modeling of nerve damage requires prolonged elevation of IP to evaluate optic nerve head susceptibility to IP and how it is affected by age, systemic diseases, drugs, and other factors. Nonhuman primate animal models 910 are expensive, which limits their usefulness for detailed study of chronic glaucomatous optic nerve damage. ther chronic models in dogs 11 are similarly expensive, and those described in rabbits 12 suffer from marked differences in optic nerve head structures. We therefore sought to develop such a model in rats because they Investigative phthalmology & Visual Science, February 1993, Vol. 34, No. 2 Copyright. Association for Research in Vision and phthalmology 363

2 364 Investigative phthalmology 8c Visual Science, February 1993, Vol. 34, No. 2 are less expensive and are easier to handle. Furthermore, a large body of literature on the cell biology of rat optic nerve damage from various causes already exists. 13 A successful chronic model of pressure-induced optic nerve damage, however, requires accurate, reproducible, and noninvasive measurement of IP over a period of months. Such requirements make the Tono-Pen 2 (BioRad, Santa Ana, CA), a hand-held tonometer, especially attractive because it is compact, portable, easily calibrated, and it has been generally reported to yield acceptable results in humans. 14 " 21 This report compares the accuracy and reproducibility of the Tono-Pen 2 tonometer against a calibrated pressure transducer in the living, cannulated rat eye. MATERIALS AND METHDS All experiments were performed in compliance with the ARV Statement for the Use of Animals in phthalmic and Vision Research. Forty male brown Norway {Rattus nowegicus) retired breeder rats were anesthetized by intraperitoneal injection of 1.5 ml/kg of a solution of 5 ml ketamine (100 mg/ml), 2.5 ml xylazine (20 mg/ml), 1 ml acepromazine (10 mg/ml), and 1.5 ml sterile water. A drop of 0.5% proparacaine hydrochloride was applied to each eye. ne cornea of each rat was punctured nasally with a 23 G hypodermic needle. A cannula made from the blunted tip of a 23 G needle connected to a length of polyethylene tubing (PE-50; Clay Adams, Parsippany, NJ) was inserted into the anterior chamber. The opposite end of the PE-50 tubing was connected to a 23 G needle with the hub connected simultaneously to a disposable "mikro-tip" catheter pressure transducer (model no. MPC 500; Millar Instruments, Houston, TX) and a 1-ml tuberculin syringe by way of a three-way stopcock for manipulating IP. Intraocular pressure was varied and maintained at each pressure interval by injecting balanced salt solution (BSS) from the syringe into the anterior chamber. This closed system was filled with BSS before cannulation, and all air bubbles eliminated. A continuous readout of IP measured by the transducer was recorded by a chart recorder (Gould RS 3800, Cleveland, H), calibrated to the transducer using the pressure gauge from a blood pressure cuff. After cannula insertion, light pressure on the eye produced immediate chart recorder pen deflection, demonstrating the patency of the system and the validity of chart recorder readings as a reflection of IP. Intraocular pressure was measured with a Tono- Pen 2, upgraded by the manufacturer to the specifications of a Tono-Pen XL. With the eye under good illumination, the Tono-Pen was oriented perpendicular to the cornea and, using a swift and steady stroke, the tip was brought into contact with the cornea. All valid readings occurred immediately after contact of the Tono-Pen tip with the eye, holding the contact long enough to see the pressure readout. The Tono- Pen would occasionally report a pressure just before the tip contacted the cornea (contact with the tear film meniscus), or just after the tip was removed from the cornea. Such readings were found to be unreliable and were systematically ignored, as were the periodic internally averaged readings. In this manner, 15 valid readings at the set IP level were recorded. Then, IP was changed to the next pressure level and 15 more Tono-Pen readings were collected in identical fashion. The rats were divided evenly into two groups group 1 and group 2. f the group 1 rats, 12 right eyes and 8 left eyes were cannulated. Intraocular pressure was raised in 5-mmHg increments, from 15 to 45 mnihg. To perform these measurements, one experimenter cannulated the eye and measured IP with the Tono-Pen, and the other controlled IP with the syringe and recorded valid pressure readings. The same two experimenters performed all of the measurements, alternating between measuring with the Tono- Pen and varying the IP so that each experimenter performed each function 10 times. The eye was periodically irrigated with balanced salt solution, and the excess was blotted. f the group 2 rats, 10 right eyes and 10 left eyes were cannulated. These experiments were performed in the same manner as in group 1, except the pressures were varied randomly in 5- or 10-mmHg increments in the 15- to 35-mmHg range. These readings were performed in a masked fashion, so that the person using the Tono-Pen did not know what the actual IP was. The pressure levels were chosen by the experimenter monitoring the IP, but with the stipulation that pressures be limited to 35 inmhg maximum and that there be no more than a 10-mmHg difference between consecutive pressures to minimize acute effects of repeated large IP fluctuations on the thin-walled cornea. For each eye, the mean of the 15 valid readings per pressure level was calculated, producing 7 mean pressures per rat in group 1 and 5 per rat in group 2. This yielded 20 data points at each pressure level for each group. Data for group 1 and group 2 were analyzed separately using linear regression analysis, 22 correlation coefficients, student t-test, one- and twoway analysis of variance, 22 and a method comparison analysis. 23 RESULTS The correlation of the Tono-Pen with the transducer was high for both group 1 (Fig. 1) and group 2 (Fig. 2), as shown by linear regression. The assumption of lin-

3 Measurement of Rat IP With the TonoPen 365 en S3 E i B a. o i Line of Equality Regression Line TABLE l. Mean Difference Between Group 1 Tono-Pen and Transducer Measurements Transducer Pressure (mmhg) Mean Difference* * Transducer minus Tono-Pen. t Student's (-test. SD ±1.1 1 ±1.21 ±1.25 ±1.07 ±1.38 ±2.28 ±1.62 Pf Transducer Pressure (mm Hg) FIGURE l. Scatter diagram for group 1 data (n = 140). Regression line formula is y = x (r = 0.98). For the slope of the line, t = (P < 0); s y. x = earity of the data was supported by analysis of variance, yielding P < 0 for both groups. For group 1, the standard deviation of the observations about the fitted regression line, s y. x, was 1.51 mnihg, whereas for group 2, s y. x was 2.08 mmhg. These standard deviations were small, but suggested that the group 1 measurements were slightly more consistent, overall. X B v a o o H C ~ ~ Line of Equality Regression Line Transducer Pressure (mm Hg) FIGURE 2. Scatter diagram for group 2 data (n = 100). Regression line formula is y = x (r = 0.94). For the slope of the line, t = (P < 0); s y. x = In both groups, the Tono-Pen overestimated transducer pressure in the 15- to 20-mmHg range and it underestimated pressures at IP of 25 mmhg and above (Tables 1, 2). This was supported by Student's t-test and a method comparison analysis that compared, at each pressure level, the mean of the difference between transducer pressures and Tono-Pen readings (transducer minus averaged Tono-Pen) with the average pressure from the two methods. This bias was significant at extremes of pressure, shown by the slopes of the lines and the corresponding high correlation coefficients, as illustrated in Figures 3 and 4. To determine whether there was a significant difference between the group 1 and group 2 data, a twoway analysis of variance was performed. The difference between groups 1 and 2 was not significant (P > 0.76). The difference among Tono-Pen readings when grouped by pressure level was highly significant (P < 0), and different combinations of pressure levels and experiment groups produced no significant effects (ie, no significant interaction; P ^ 0.47). Because Tono-Pen measurements differ significantly, but predictably from transducer measurements, we adopted the group 1 regression line as a correction formula for the Tono-Pen data. The calculated (corrected) IP from hypothetical mean Tono- TABLE 2. Mean Difference Between Group 2 Tono-Pen and Transducer Measurements Transducer Pressure (mmhg) Mean Difference* * Transducer minus Tono-Pen. t Student's t-tesl. SD ±1.12 ±1.51 ±2.12 ±2.10 ±3.06 Pf

4 366 Investigative phthalmology & Visual Science, February 1993, Vol. 34, No. 2 CUD X E 15 -i 5- e 4) Q. H u Regression Line Average Pressure by Two Methods FIGURE 3. Difference (transducer minus Tono-Pen) against the mean (average of transducer and Tono-Pen) for group 1 data. Regression line formula is y = x (r = 0.77). For the slope of the line, t = (P < 0); s y. x = The mean is 1.90 mmhg, and 2 SD is ± Pen values, with 95% prediction intervals, 24 are shown in Table 3 and Figure 5. DISCUSSIN Previous studies using human subjects to compare the Tono-Pen against manometric intraocular pressure or X) X B 3 5- Mean + 2SD ft, Q. H u 4) U T3 CB o- against Goldmann tonometry over a wide range of pressures have indicated that the Tono-Pen is quite accurate. 16 ' 25 ' 26 Certain variables inherent in a living system were eliminated from these studies, however, because they used gas-filled, postmortem, or enucleated eyes. In studies using living subjects, some investigators have found the Tono-Pen to provide an accu Mean Mean - 2SD Regression Line Average Pressure by Two Methods FIGURE 4. Difference (transducer minus Tono-Pen) against the mean (average of transducer and Tono-Pen) for group 2 data. Regression line formula is y = x (r = 0.46). For the slope of the line, t = (P < 0); s y. x = The mean is 0.70 mmhg, and 2 SD is ± 5.18.

5 Measurement of Rat IP With the TonoPen 367 TABLE 3. Tono-Pen Pressures Calculated Using the Correction Formula Hypothetical Tono-Pen Reading Calculated Pressure (mmhg) % Prediction Intervals ±3.80 ±3.80 rate measure of IP, whereas others have concluded the Tono-Pen is unacceptable for clinical use. 19 " 21 Because IP fluctuates significantly over short periods of time, such as with pulse, any tonometer that records a near-instantaneous measurement of IP is really only taking a sample from the IP cycle, which varies about the true IP. So, variables such as fluctuating blood pressure, pulse, respiration, and anxiety could account for some discrepancies in the accuracy of Tono-Pen measurements, 1627 " 29 as could the stress of repeated applanations. 30 Although not all of the above variables were controlled for in our experiments, our results suggest that the Tono-Pen 2 tonometer provides a reliable method of noninvasively measuring IP in the living rat eye. We collected 105 Tono-Pen measurements for each group 1 rat, and 75 measurements for each group 2 rat, for a total of 3600 measurements. It was these data that were averaged by pressure level, for each rat, to obtain our results. A linear relationship between Tono-Pen and transducer values was established by analysis of variance, showing that both the transducer and the Tono-Pen measure IP. By two-way analysis of variance, the results for groups 1 and 2 are shown to be remarkably similar. This indicates the repeatability and consistency of the measurements between the two groups, one of them performed in random, masked fashion. Although overestimation by the Tono-Pen in rat eyes at low IP and underestimation at high IP was statistically significant, such overestimation and underestimation has been encountered in experiments comparing the Tono-Pen 2 against other tonometers in human subjects Because the differences between Tono-Pen and transducer readings were consistent between groups 1 and 2, a correction formula with 95% prediction intervals can be generated and applied (Fig. 5 and Table 3). The prediction interval of 3.8 mmhg, as indicated in Table 3, is comparable to the error associated with other methods, such as the Goldmann tonometer, where the standard 9) Q. I deviation on human subjects has been reported to be about ± 3 mmhg 30 and is at least ± 2 mmhg. 31 Even without applying a correction formula, the average Tono-Pen readings reported here agree surprisingly well with actual IP, despite the marked differences in the corneal thickness and radius of curvature between rat eyes and human eyes, against which the Tono-Pen is calibrated by the manufacturer. Such agreement is encouraging for the potential use of this instrument in measuring IP in a future model of rat glaucoma, in which chronically elevated IP potentially may produce enlargement of the cornea and subsequent flattening of its radius of curvature. Successful use of the Tono-Pen on the relatively small rat eye depends on a number of factors. Using good illumination of the eye, the Tono-Pen tip must be oriented exactly perpendicular to the cornea. This maximizes the chance that the central post of the Tono-Pen tip will have full contact with the highly curved cornea with each applanation. The tonometer must be applied using a swift and steady stroke, stopping after bringing the tip into contact with the eye, but avoiding excessive pressure. An unmodified Tono- Pen 2 or Tono-Pen XL is essential. These "secondgeneration" instruments display a reading every time a postmovement waveform found acceptable to the instrument is encountered, rather than providing a statistical average only if several such readings occur within a specified time period. It is these individual readings that were averaged in these experiments, igy = x R = Transducer Pressure (mm Hg) FIGURE 5. Scatter diagram for group 1 data, with 95% prediction intervals shown. The prediction interval is roughly 3.81 mmhg throughout the 15- to 45-mmHg range (along the y axis).

6 368 Investigative phthalmology & Visual Science, February 1993, Vol. 34, No. 2 noring the instrument-generated averages that we found were inaccurate and too infrequent to be of practical use. Finally, we are able to discriminate valid from invalid readings on a systematic basis. Valid readings are those produced immediately after firm contact with the cornea ("on" readings). Contact must be held long enough to allow the observer to note the pressure reading, since a second reading may be generated when the tip breaks contact with the cornea ("off" reading). Such "off" readings have proved unreliable, possibly because of the alteration of IP by the instrument itself. ther invalid readings occur when the tip lightly touches the cornea, producing an artifactually low reading, probably from influence of the tear-film meniscus. We also have found that abrupt movements of the Tono-Pen without corneal contact occasionally can produce a spurious reading. A statistical analysis of our data to determine the required sample size for estimating means indicated that at lower pressures, 10 readings would be sufficient to obtain an acceptable standard deviation, whereas at pressures over 30 mmhg, 15 measurements are necessary. Hence, when using the instrument to measure IP noninvasively, 15 valid measurements would be required. In our experience, this number of readings can be collected in less than 2 min per eye. We have found an immediate feedback system, such as the transducer arrangement described here, is helpful but not necessary to master and maintain operator skill. Key Words intraocular pressure, rat, tonometry, Tono-Pen. Acknowledgments The authors thank Dr. Jiten Vora for his invaluable assistance with the statistical analysis of our data. References 1. Tielsch JM, Sommer A, Katz J, Royall RM, Quigley HA, Javitt J. Racial variations in the prevalence of primary open-angle glaucoma. JAMA. 1991; 266: Zeimer RC, gura Y. The relation between glaucomalous damage and optic nerve head mechanical compliance. Arch phthalmol. 1989; 107: Quigley HA, Addicks EM. Regional differences in the structure of the lamina cribrosa and their relation to glaucomatous optic nerve damage. Arch phthalmol. 1981;99: Sossi N, Anderson DR. Blockage of axonal transport in optic nerve induced by elevation of intraocular pressure. Effect of arterial hypertension induced by Angiotensin I. Arch phthalmol. 1983; 101: Johansson J. Retrograde axoplasmic transport in rat optic nerves in vivo. What causes blockage at increased intraocular pressure? Exp Eye Res. 1986;43: Johansson J. Inhibition and recovery of retrograde axoplasmic transport in rat optic nerve during and after elevated IP in vivo. Exp Eye Res. 1988;46: de Kater AW, Smyth JR Jr, Rosenquist RC, Epstein DL. The Slate turkey: A model for secondary angle closure glaucoma. Invest phthalmol Vis Sci. 1986; 27: Radius RL, Anderson DR. Rapid axonal transport in primate optic nerve. Distribution of pressure-induced interruption. Arch phthalmol. 1981;99: Quigley HA, Hohman RM. Laser energy levels for trabecular meshwork damage in the primate eye. Invest phthalmol Vis Sci. 1983;24: Quigley HA, Addicks EM. Chronic experimental glaucoma in primates. I. Production of elevated intraocular pressure by anterior chamber injection of autologous red blood cells. Invest phthalmol Vis Sci. 1980;19: Samuelson DA, Gum GG, Gelatt KN. Ultrastructural changes in the aqueous outflow apparatus of beagles with inherited glaucoma. Invest phthalmol Vis Sci. 1989;30: Bunt-Milam AH, Dennis MB Jr., Bensinger RE. ptic nerve head axonal transport in rabbits with hereditary glaucoma. Exp Eye Res. 1987;44:537-55L 13. Doster SK, Lozano AM, Aguayo AJ, Willard MB. Expression of the growth-associated protein GAP-43 in adult rat retinal ganglion cells following axon injury. Neuron. 1991; 4: Minckler DS, Baerveldt G, Heuer DK, Quillen- Thomas B, Walonker AF, Weiner J. Clinical evaluation of the culab Tono-Pen. Am J phthalmol. 1987;104: Kao SF, Lichter PR, Bergstrom TJ, Rowe S, Musch DC. Clinical comparison of the culab Tono-Pen to the Goldmann applanation tonometer. phthalmology. 1987;94: Boothe WA, Lee DA, Panek WC, Pettit TH. The Tono-Pen. A manometric and clinical study. Arch phthalmol. 1988; 106: Kooner K, CookseyJ, Barron J, Zimmerman T. Goldmann tonometry vs. culab TonoPen. Invest phthalmol Vis Sci. 1988;29(suppl): Khan JA, Davis M, Graham CE, Trank J, Whitacre MM. Comparison of culab Tono-Pen readings obtained from various corneal and scleral locations. Arch phthalmol. 1991; 109: Armstrong TA. Evaluation of the Tono-Pen and the Pulsair tonometers. Am J phthalmol. 1990; 109: Farrar SM, Miller KN, Shields MB, Stoup CM. An evaluation of the Tono-Pen for the measurement of diurnal intraocular pressure. Am] phthalmol. 1989; 107: Higginbotham EJ. Clinical evaluation of the culab Tono-Pen. Am J phthalmol. 1988; 105: Daniel WW. Biostatistics: A Foundation for Analysis in the Health Sciences. 5th ed. New York: Wiley; 1991.

7 Measurement of Rat IP With the TonoPen Bland JM, Altman DC Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1987;2: Snedecor GW, Cochran WG. Statistical Methods. 8th ed. Iowa: Iowa State University Press Hessemer V, Rossler R, Jacobi KW. Comparison of intraocular pressure measurements with the culab Tono-Pen vs manometry in humans shortly after death. AmJ phthalmol. 1988; 105: Hines MW, Jost BF, Fogelman KL. culab Tono-Pen, Goldmann applanation tononietry, and pneumatic tonometry for intraocular pressure assessment in gasfilled eyes. AmJ phthalmol. 1988; 106: Moses RA, Ching-Hung L. Repeated applanation tonometry. AmJ phthalmol. 1968;66: Moses RA, Arnzen RJ. Instantaneous tononietry. Arch phthalmol. 1983; 101: Lichter PR, Bergstrom TJ. Premature ventricular systole detection by applanation tononietry. AmJ phthalmol. 1976;81: Motolko MA, Feldman F, Hyde M, Hudy D. Sources of variability on the results of applanation tononietry. CanJ phthalmol. 1982; 17: Minckler DS, Baerveldt G, Heuer DK, Quillen- Thomas B, Walonker AF, Weiner J. Clinical evaluation of the culab Tono-Pen (Reply). AmJ phthalmol. 1988; 105:101.