Far ultraviolet absorption spectra ofporcine and human comeas Annamarie Lembares', Xin-Hua Hu', Gerhard W. Kalmus2 (1) Biomedical Laser Laboratory, Department of Physics (2) Embryology Histology Laboratory, Department of Biology East Carolina University, Greenville, NC 27858 ABSTRACT We have measured the absorption spectra ofporcine and human comeas in the far ultraviolet region from 350 nm to 190 nm. The samples were sectioned from frozen comeas into thin slices of 15 tm to 40.tm in thickness. The absorption spectra shows three distinct regions from 260 nm to 190 nm: a relatively weak absorption region from 260 nm to 240 nm; a region of steep increase from 240 nm to 220 nm; and a strong and slowly increasing absorption region from 220 nm to 190 nm. The linear absorption coefficients were determined to be 2130 3 10 (cm') at 220 nm, 2240 320 (cm') at 215 nm, 2300 330 (cm1) at 210 nm, and 2410 370 (cm') at 193 nm for porcine corneas and 2300 480 (cm') at 220 nm, 2300 460 (cm') at 215 nm, 2320 470 (cm') at 210 nm, and 2340 450 (cm') at 193 nm for human corneas. A "window of ablation" in the far ultraviolet region between 220 nm and 190 nm was determined. Statistical analysis was conducted to correlate the far ultraviolet absorption between the porcine and human corneas. The effect offreezing on the far ultraviolet absorption also has been studied. Key Words: absorption spectra, cornea, comeal ablation, comeal absorption, corneal surgery, far UV absorption 1. INTRODUCTION Solid state laser systems are being investigated as a possible replacement to the excimer (ArF) laser systems for treating refractive errors ofthe eye through corneal ablation. The solid state laser systems can provide short ultraviolet (UV) pulses through harmonic generation at wavelengths near 210 nm.12 It is well known that the cornea strongly absorbs far UV radiation at 193 nm which makes the ArF excimer lasers the current choice for precise ablation of the cornea. Hence, a significant amount of data has been collected on the corneal ablation at 193 nm for the excimer laser systems.35 However, the measurement ofthe corneal absorption between 190 nm to 230 nm has not been reported.6 Thus, it is difficult to directly correlate corneal ablation with solid state laser pulses near 210 nm to the results ofarf lasers at 193 nm since the ablation process strongly depends on the far UV absorption. An earlier investigation of corneal ablation with picosecond laser pulses at 2 1 1 nm and 263 nm suggested that the corneal absorption in the far UV region may be divided into three segments based on the analysis 2 of collateral tissue damage I was concluded that the corneal absorption is relatively weak from 266 nm to 248 nm, increases steeply from 248 nm to 213 nm, and remains strong from 213 nm to 193 nm.2 In this report, we present a study ofthe far UV absorption spectrum ofthe porcine and human corneas from 350amto 190 mn and its dependence on tissue conditions. We also have performed a statistical analysis to correlate the far ultraviolet absorption spectra between the porcine and human corneas at wavelengths of 220 nm, 215 nm, 210 mn, and 193nm. 2. METHODS 2.1 Samples Porcine and human corneas were used for the absorbance measurements following institutional guidelines regarding the use oftissue and organs. Fresh porcine eye globes were obtained from a local slaughter house and the School of Medicine, East Carolina University. A total of 23 corneal samples sectioned from 1 8 porcine corneas and 3 whole porcine corneas were used for the study. All eye globes were removed from the animal immediately after death and stored on ice until the corneas were excised from the globes. Absorbance measurements of porcine corneas were performed within 12 hours postmortem except those samples used for studying the effect of freezing 46 SPIE Vol. 2971 0277-786X/97/$1O.OO
time. A total of 1 1 corneal samples sectioned from 5 fresh human corneas and 1 preserved human cornea were also used. Both the human eye globes and the preserved human cornea were provided by the North Carolina Eye and Human Tissue Bank. The absorbance ofthe fresh human corneas were measured within 33 hours postmortem and the preserved human cornea at 1 week postmortem. 2.2 Sectioning ofthe Cornea Since the cornea strongly absorbs far UV radiation, the sample thickness needs to be in the vicinity of 20 tm to keep the absorbance within the scale ofthe spectrophotometer. To obtain a thin and contiguous corneal sample of uniform thickness, a microtome-cryostat (Model 455 1, Ames Company) was used to section a frozen cornea at -1 8 C. The corneas were excised from refrigerated eye globes, stored in saline and frozen in a freezing preservation medium for at least 15 minutes before sectioning. The superficial layers ofthe cornea, including the epithelium, were removed until a uniform slice from the corneal stroma was obtained. The slice was transferred to a sample holder comprised oftwo rectangular optical windows made ofuv grade fused silica (Type 7940, Corning Inc.) which has a very low thermal expansion coefficient that keeps the structure ofthe thin corneal sample intact in the course ofthawing and transferring to the spectrophotometer. The sample holder has a total thickness of 12 mm and a very low absorbance to help maintain the total absorbance within the scale of the spectrophotometer. The absorbance ofthe sample holder increases slowly as the wavelength decreases in the far UV region, which was measured to be less than 0.01 at 260 rim and to be 0.01 1 at 190 nm after the surface reflection contribution was deducted. Using a micrometer of 3 tm (0.0001 inch) resolution, the thickness of a corneal sample was determined through the measurement ofthe thickness difference ofthe sample holder with and without the corneal sample at room temperature. The ratchet stop mechanism on the micrometer enabled a consistent application of measuring pressure on the tissue for each measurement. Furthermore, the thickness measured by the micrometer was verified by the setting ofthe knife advancement in the microtome. The estimated uncertainty in the thickness measurement is less than 20%. 2.3 Absorbance Measurement The absorbance measurements ofthe corneal samples were measured at room temperature with a dual-beam UV-Visible-IR spectrophotometer (Model 17D, Varian Associates) in the UV region from 350 nm to 190 nm. The total absorbance can be measured from 0.0 to 3.0 in 5 scales by the spectrophotometer for wavelengths between 3000 nm and 186 nm. A standard UV mercury lamp was used to calibrate the UV wavelength reading and a dye solution with known absorbance for the absorbance reading. Both were found to be within manufacturer's specifications. To collimate the incident and transmitted light beams, two apertures of4 mm in diameter were used at the front and back surfaces ofthe sample holder. The output signal from the spectrophotometer was digitized and averaged by a personal computer with an A/D board. The total absorbance of the sample and the sample holder in the sample chamber was measured and the corneal absorbance was obtained by subtracting the absorbance ofthe sample holder from the total absorbance. 3. RESULTS The absorbance oftwo sectioned samples oftwo porcine corneas as a function ofwavelength from 260 mm to 190 mm is shown in Figure 1. In the insert of Figure 1, one full spectrum ofthe absorbance is plotted from 350 mm to 190 rim. The absorbance ofhuman corneal samples as a function ofwavelength from 260 urn to 190 mm has also been measured and is shown in Figure 2, with one full spectrum from 350 mm to 190 mm displayed in the insert. One ofthe human corneal samples in Figure 2 was placed in a preservation solution (Optisol, Chiron Vision) for 1 week before the measurement and exhibited a light pink color after it was removed from the preservation cell. Figures 1 and 2 show a larger fluctuation in the absorbance readings in the short wavelength region below 195 mm than that in the longer wavelength region. This was expected due to the relative large noise presented in a weak signal which was caused by the strong absorption of the corneal sample and the low light intensity near the short wavelength end of the emission spectrum of the light source. Our measurements indicate that the dependency of the corneal absorption on wavelength is very consistent between 260 mm to 190 mm for both porcine and human corneas even though the absorbance fluctuates from sample 47
to sample with similar thickness. All the absorbance spectra show that from 240 nm to 220 nm the corneal absorption increases significantly when the wavelength decreases. Based on this steep increase, the cornea! absorption in the far UV region between 260 nm and 190 nm can be represented by three segments with clear boundaries at 240 nm and 220 urn: a weak absorption segment between 260 nm and 240 nm, a steeply increasing segment between 240 nm and 220 nm, and a strong slowly increasing absorption segment below 220 nm. Since the scattering ofthe UV radiation by the cornea is significantly less than the absorption ofthe UV radiation, we can use the Lambert's law to fmd the transmittance T of a corneal sample as: T=I/10=(1 R)2e, (1) where I and I are the light intensity at the front and rear surfaces ofthe sample holder, R is the Fresne! reflectivity ofthe glass-cornea interfaces, a is the linear absorption coefficient ofthe cornea and d is the thickness of the cornea! sample. The total absorbance ofthe sample and the sample holder in the sample chamber was measured through the dual-beam configuration ofthe spectrophotometer. The cornea! absorbance A was obtained by subtracting the absorbance ofthe sample holder from the total absorbance. Therefore, A is related to T by: A = 1og10(1/T) = O.434ad 2. log10(l R). (2A) If 2. 1og10(1 R) << O.434ad, then Equation (2A) becomes A O.434ad. (2B) However, the reflection loss contribution to the measured absorbance can not be calculated since the refraction index ofthe cornea in the far UV region is not known. But we can estimate that R is smaller than 0.1 in this spectral region from the refraction index of water. Thus, Equation (2B) should only be used to accurately calculate the linear absorption coefficient a from the measured A in the segment of strong absorption between 220 nm and 190 rim in which the absorbance is well above 1.0. Table 1 shows the results ofthickness measurements and the calculated a ofporcine corneas, from Equation (2B), at four wavelengths of220 nm, 215nm, 2lOnm, and 193nm. The linear absorption coefficients a were determined to be 2130 310 (cm1) at 220 nm, 2240 320 (cm1) at 215 urn, 2300 330 (cm1) at 210 nm, and 2410 370 (cm') at 193 nm from 23 samples out of 18 porcine corneas. Table 2 lists similar results for human corneas where a was calculated to be 2300 480 (cm1) at 220 nm, 2300 460 (cm') at 215 nm, 2320 470 (cm') at 210 nm, and 2340 450 (cm') at 193 urn from 1 1 samples out of6 human corneas. The standard deviations in the absorption coefficients for porcine corneas are consistent with our expected dominant source of error in the thickness measurement. The correlation between the porcine and human linear absorption coefficients at 220 nm, 2 1 5 urn, 2 10 nm and 193 nm were analyzed by the unpaired Student's t-test using the two-tailed approach with the null hypothesis and a confidence level of 95%. The calculated tratio was less than the critical value of tatthe 95% confidence level for the four wavelengths. This indicates that the null hypothesis can not be rejected. Therefore, the linear absorption coefficients ofthe porcine and human corneas are less than 5% significantly different at each wavelength. The effect of freezing on the absorbance measurement was investigated through two sets of tests with porcine cornea! samples. The absorbance of3 whole corneas from 350 nm to near 290 nm was measured before and after freezing. Figure 3 shows the results from one ofthe absorbance measurements where the absorbance spectra differ in the long wavelength region but converge toward 290 nm when the absorbance readings approach the maximum of the scale. The effect of freezing time on the absorbance of porcine cornea! sample was also studied with the variation in freezing time from 15 minutes to 10 hours, which is shown in Table 1. There was no significant change found due to the length of freezing time in the spectral region from 260 nm to 190 nm. Because of the technical difficulty in quality sectioning fresh corneas, we were unable to directly measure the effect of freezing on cornea! absorbance in the far UV region. 48
4. DISCUSSION To our best knowledge, this is the first time a study ofthe complete spectrum ofthe far UV absorbance spectrum ofthe cornea between 260nm and 190 nm has been reported. The results show that the corneal absorption between 260 nm and 190 nm can be divided into three segments. Also, the average value ofthe linear absorption coefficient ofthe corneal samples remains approximately the same as the wavelength decreases from 220 nm to 190 nm. These conclusions are consistent with a previous tu2 on the corneal absorption based on the analysis of collateral tissue damage in the corneal ablation with short laser pulses in the far UV region. In examining our measurement ofthe linear absorption coefficient ofthe cornea at 193 urn, we found that it is about 15% less than that measured by Puliafito et al..5 Thickness measurements taken of 3 whole porcine corneas showed a decrease in the average thickness from 8 10 tm before the freezing to 720 im after the freezing. This may be attributed to dehydration in the freezing and thawing processes. If a similar reduction in thickness occurred in the sectioned samples, our determination ofthe linear absorption coefficient from the measured absorbance may overestimate the coefficient by about 10%. In addition to the effects of freezing on the absorbance spectra, the porcine eyes obtained from the slaughterhouse were removed from scalded pigs while the porcine eyes used for the last five samples in Table 1 were obtained from the School of Medicine without scalding. We have found no significant effect due to scalding in our measurement ofthe far UV absorbance ofporcine corneas. The effect of freezing on the absorbance measurement has not been directly studied in the far UV region between 260 urn and 190 nm. Therefore, some uncertainty exists in correlating our results to the clinical investigations of corneal ablation. Previous studies indicated that among the major corneal components only collagen has shown a steep rise in the UV absorption spectrum below 240 nm, which is similar to our results reported here. Since small amino acids linked in long molecular chains by peptide bonds compose collagen and freezing involves only the phase transition between water and ice, it is reasonable to assume that freezing has a minimal effect on corneal collagen. Thus, any change in corneal absorption due to freezing may be insignificant compared to the major source of uncertainty in determination of sample thickness. It has been suggested that the size of collateral tissue damage zones is primarily dependent on the linear absorption coefficient ofthe cornea.2 Our results on corneal absorption provided an indirect confirmation of this hypothesis. Furthermore, we may conclude from this report that laser pulses with wavelengths in a "window of ablation" between 220 nm and 190 urn can be used for the surface ablation ofthe cornea with comparable outcomes. 5. REFERENCES 1 Ren Q, Gailitis RP, Thompson KP, Lin JT, "Ablation ofthe cornea and synthetic polymers using a UV (2 13 nm) solid state laser", IEEE.1 Quan. Electron., 26:2284-2288, 1990 2 Hu XH, Juhasz T, "Study ofcorneal ablation with picosecond laser pulses at 21 mm and 263nm'ç Lasers Surg. Med, 18:373-380, 1996 3 Trokel SL, Srinivasan R, Braren BB, "Excimer laser surgery ofthe cornea", Am..1 Ophthalmol., 96:710-714, 1983 4 Krueger RR, Trokel SL, Schubert HD, "Interaction ofultraviolet laser light with the cornea", mv. Ophthalmol. Vis. Sci., 26:1455-1464, 1985 5 Puliafito CA, Steinert RF, Deutsch TF, Hillenkamp F, Dehm EJ, Adler CM, "Excimer laser ablation of the cornea and lens", Ophthalmology, 92:741-748, 1987 6 Mitchell J, Cenedella RI, "Quantitation of ultraviolet light-absorbing fractions of the cornea", Cornea, 14:266-272, 1995 7 Loolbourow JR, Gould BS, Sizer 1W, "Studies on the ultraviolet absorption spectra of collagen", Arch. Biochem., 22:406-411, 1949 49
3.0 2.5 2.0 C-) Cl) -Q 1.5 1.0 0.5 0.0 190 200 210 220 230 240 250 Wavelength (nm) 260 Figure 1 The absorbance dependence on the wavelength for two porcine corneal samples between 260 nm and 190 nm. Insert: The absorbance dependence on the wavelength for one porcine corneal sample between 350nm and l9onm. 50
3.0 ci C-) C a) 2 2.5.0 0Cl).0 I 2.0 -,_t- ' 0 200 250 300 350 Wavelength (nm) C-) Co.0 1.5 \ \\ Cl).0 1.0 thickness: 19 im (7061A/preserved) thickness: 21 jtm (8192A) 0.5 0.0 190 200 210 220 230 240 250 260 Wavelength (nm) Figure 2 The absorbance dependence on the wavelength for two human cornea! samples between 260 nm and 190 rim. Insert: The absorbance dependence on the wavelength for one human cornea! sample between 350 nmandl9onm. 51
3.0 before freezing afterfreezing 2.5 2.0 1.5 Cl) \ \ 1.0 \ \\ 0.5 0.0 I t!iii!tii I tt!!t lilt I ititittit 111111 lit 280 290 300 310 320 330 340 350 Wavelength (nm) Figure 3 The absorbance dependence of one whole porcine cornea before and after freezing. 52
Table 1 : The thickness and linear absorption coefficients of porcine cornea samples sample # thickness a (cm') a (cm') a (cm') a (cm') postmortem* / freezing time (tm) at 220 nm at 215 nm at 2 10 nm at 193 nm (hour) 6271 lb 20 2230 2290 2350 2430 1.5 I 1.2 62711c 23 1820 1940 2010 2120 1.5/9.0 62712a 21 1820 1930 1990 2120 1.3/1.5 62731b 25 1700 1760 1790 2080 3.3/3.2 71121a 26 2490 2510 2590 2290 5.2/0.6 71122c 19 2680 2880 2940 3010 7.2/5.6 71131a 27 1850 1910 1970 1940 7.2/7.5 72911a 18 2030 2180 2340 2510 1.5/0.3 72911b 18 2330 2410 2520 2670 1.5/10.5 72912a 26 2130 2290 2350 2630 2.4/0.4 72912b 23 2360 2520 2580 2680 2.4/1.2 72923a 21 2030 2140 2140 2260 5.2/0.7 72936a 12 2630 2710 2710 2630 9.0/0.7 81923a 15 2430 2640 2700 3180 5.6 / 1.0 81912b 21 2450 2620 2750 2910 2.8/2.8 81923b 20 2070 2330 2480 2800 5.2/2.0 81924b 21 1800 1900 1930 2100 7.6/5.0 81924c 24 1910 2030 2130 2190 7.6/5.9 ilolla 30 2090 2170 2190 2160 2.0/0.5 11021b 21 1610 1680 1710 1660 4.3/1.5 11022a 15 2570 2590 2600 2510 4.3/2.7 1 1031a 16 1960 2000 2020 2220 6.8 / 1.2 11032a 28 2090 2180 2230 2280 6.8/2.2 * The postmortem time is defmed to be the period between animal death and the beginning of corneal freezing. 53
Table 2: The thickness and linear absorption coefficients ofhuman cornea samples sample # thickness a (cm') a (cm') a (cm') a (cm') postmortem time (tm) at220 nm at 21 5 nm at 2 10 nm at 193 nm 7061a 19 2590 2680 2750 2920 1 week @reserved) 7061b 28 2080 2090 2090 2180 lweek@reserved) 7062a 15 1970 2030 2030 2 120 33 hours 7062b 20 1600 1660 1670 1720 33 hours 7062c 15 3000 3180 3210 3140 33 hours 8191a 17 2710 2740 2760 2680 28 hours 8192a 21 2920 2290 2250 2120 30 hours 0301a 21 2450 2530 2540 2420 28 hours 0301b 24 1650 1680 1680 1680 28 hours 0302a 21 2090 2160 2220 2390 28 hours 0302b 28 2260 2260 2290 2380 28 hours 54