Effects of Oxygen and Carbon Dioxide on Human Retinal Circulation

Effects of ygen and on Dioxide on Human Retinal Circulation Stephen]. Pakola and Juan E. Grunwald Purpose. ogen, a gas mixture of 95% O 2 and 5% CO 2, is given to patients with retinal artery obstruction in an attempt to improve retinal oxygenation. The purpose of this study was to compare the effects of carbogen and 100% O 2 breathing on retinal blood flow. Methods. On two separate occasions, 12 normal, healthy volunteers breathed air and then either 100% O 2 or carbogen while laser Doppler velocimetry measurements and monochromatic fundus photographs were taken. Retinal vessel diameter, maximum velocity of red blood cells, and volumetric blood flow rate were determined in a main temporal vein. Results. Both 100% O 2 and carbogen caused significant average reductions in vessel diameter (14.1% and 10.6%, respectively), maximum red blood cell velocity (42.1% and 27.3%, respectively), and blood flow (56.4% and 42.2%, respectively). The average vasoconstriction of the large retinal veins caused by carbogen was not significantly smaller than that caused by 100% O 2. The average reductions in maximum red blood cell velocity and blood flow caused by carbogen were significantly smaller than those caused by 100% O 2 (P <.001 and P <.01, respectively). Conclusions. In normal subjects, inhalation of carbogen leads to less reduction in blood flow than inhalation of 100% O 2, presumably by reducing the vasoconstriction of small arterioles induced by elevated oxygen levels. Invest Ophthalmol Vis Sci 1993;34:2866-2870. Central retinal artery occlusion is a medical emergency treated by ocular massage, reduction of intraocular pressure (IOP), and inhalation therapy with gases containing high oxygen concentrations. ogen, a gas mixture of 95% O 2 and 5% CO 2, is one of the gases that has been recommended for administration 10 minutes every hour during the day and every 4 hours at night. 1 " 3 Inhalation therapy is provided with the hope that hyperoxic blood from the choroid circulation may supply the ischemic inner retina and that additional oxygen may be delivered through a partially perfused retinal circulation. Flower and Patz 4 reported that hyperbaric oxygen was necessary to improve the oxygenation of the inner retina in cats and dogs with occluded From the Scheie Eye Institute, Department of Ophthalmology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania. Supported by NIH grant EY03242. Submitted for publication ugust 7, 1992; accepted Novejnber 23, 1992. Proprietary interest category: N. Reprint requests: Juan E. Grunwald, M.D., Scheie Eye institute, 51 North 39th Street, Philadelphia, P 19104. retinal circulation. One hundred percent O 2 breathing returned the electroretinographically recorded b- wave (a measure of functional integrity of the Miiller cells in the inner retinal layers) to only 45% of normal amplitude. More recent studies, however, have suggested that normobaric 100% oxygen breathing can provide some oxygen to the inner retina in cats 5 ' 6 and miniature pigs 7 with occluded retinal circulation. In normal subjects, inhalation of 100% O 2 for 5 minutes causes a 12% and 15% decrease in the diameter of the large retinal arteries and veins, respectively, and also results in a decrease in retinal blood flow of approximately 64%. 8 Because the delivery and removal of metabolites and gases to the retina is proportional to blood flow, this large decrease in blood flow caused by 100% O 2 could perhaps have deleterious effects. on dioxide has been found to diminish the vasoconstrictive effect of high O 2 concentrations. 9 ogen is therefore used in the treatment of central retinal artery occlusion in the hope that it will cause a smaller decrease in vessel diameter and blood flow than 100% O 2. 2866 Investigative Ophthalmology & Visual Science, September 1993, Vol. 34, No. 10 Copyright ssociation for Research in Vision and Ophthalmology

Effects of O 2 and CO 2 on Human Retinal Circulation 2867 In a study of the effects of carbogen and 100% O 2 breathing for 5 minutes, Deutsch et al 10 found no significant differences between the constriction of the main retinal arteries or veins of normal subjects produced by these two gases. More recently, however, Sponsel et al," using the blue field entoptic simulation technique, found that macular capillary leukocyte velocity was significantly higher during carbogen breathing than during 100% O 2 breathing in normal volunteers. In the current study, we determined the effects of carbogen and 100% O 2 breathing not only on retinal venous diameter (D), but also on centerline maximum erythrocyte velocity (V max ), and volumetric blood flow rate (Q) using bidirectional laser Doppler velocimetry (BLDV) and monochromatic fundus photography. MTERILS ND METHODS Twelve normal, nonsmoking volunteers aged 19-33 years (mean age, 27 ± 5 years) took part in this study. ll subjects had normal results on eye examination and no history of ocular disease. None of the subjects were taking medication at the time of the study. The protocol used in this study was approved by our institutional human experimentation committee. Tenets of the Declaration of Helsinki were followed. Informed consent was obtained from all subjects after the nature of the procedure had been explained. fter pupillary dilatation with tropicamide 1%, Polaroid color fundus photographs and monochromatic fundus photographs were then taken using a Zeiss fundus camera and Polaroid Polacolor 2 (Polaroid, Cambridge, Massachusetts) and Kodak Plus-X pan (Eastman Kodak, Rochester, New York) films. Baseline BLDV measurements of V max were taken in a temporal retinal vein of one eye of each subject, first during air breathing, and then between 3V2 to 5V2 minutes of either 100% O 2 or carbogen breathing. es were administered at 1 atm through a mouthpiece while a clamp was applied to the nose. Measurements were obtained in the sitting position. One hundred percent O 2 or carbogen was given randomly without the subjects' knowledge of the type of gas. In 10 subjects, more than 24 hours passed between the administration of carbogen and 100% O 2 ; in the other two subjects, they were administered only 1 hour apart. fter the BLDV recordings and while subjects were still breathing a given gas, fundus photographs were taken. Immediately after breathing 100% O 2 or carbogen, heart rate and brachial artery systolic and diastolic blood pressure were determined using a Datascope CCUTORR 1 noninvasive blood pressure monitor (Datascope Corporation, Paramus, NJ). Topical proparacaine HCL 0.5% was applied to each eye, and IOP was determined by Goldmann applanation tonometry. Measurements of blood pressure, IOP, and heart rate are summarized in Table 1. Projected photographic negatives were used to measure the diameter of the vein at the location of the BLDV recordings. D was determined by averaging the measurements from six photographs. The method for determining V max has been described previously. 12 Retinal volumetric blood flow (Q) was calculated from the equation: Q = V me:m -7rD 2 /4 where V mean represents the mean velocity of whole blood. We have assumed that V mean = CV max, with C being a constant equal to '/i.e. 13 Measurements of D were performed by one experienced observer, and V max determinations were performed by another experienced observer. Each was masked as to the type of gas breathed. Mean brachial artery blood pressure (BP m ) was determined using the relation BP in = BP cl + 1/3(BP S - BP d ), where BP S and BP d are the brachial artery systolic and diastolic pressures, respectively. Perfusion pressure (PP) was calculated from the relation PP = 2/3BP m - IOP. Linear regression, correlation analysis, and twotailed paired Student's Mests were used in a statistical evaluation of the data. Values with a P value smaller than 0.05 were considered to be statistically significant. TBLE 1. verages and SD of Heart Rate (HR), Systolic Blood Pressure (BP S ), Diastolic Blood Pressure (BP d ), Mean Blood Pressure (BP m ), verage Intraocular Pressure (IOP), and Perfusion Pressure (PP) Immediately fter 100% ygen (O 2 ) and ogen (CO 2 ) Breathing Parameter HR, O., HR, CO, BP S, O, BP S, CO., BP fl, O, BP (1, CO, BP m, O, BP,,,, CO, IOP, O, IOP, CO, PP, O, PP, CO, verage Value 63.7 67.1 110.9 115.5 60.3 60.0 77.0 78.6 13.8 13.5 37.7 38.8 SD 7.9 6.1 11.7 10.6 11.1 7.1 10.3 7.7 1.9 2.0 5.6

2868 Investigative Ophthalmology 8c Visual Science, September 1993, Vol. 34, No. 10 RESULTS Table 2 summarizes the values of D, V max, and Q during air breathing and during 100% O 2 or carbogen breathing. Figure 1 shows the percentage decreases in D, V Iliax, and Q during O 2 or carbogen breathing, whereas Table 2 shows the absolute values of these parameters. Breathing O 2 caused significant average reductions of 14.1 ± 4.9% SD (22.3 ± 8.75 /mi) in D, 42.1 ± 12.6% (0.73 ± 0.27 cm/sec) in V max, and 56.4 ± 13.7% (7.22 ± 2.97 /*l/min) in Q (paired t-test, P <.001 in all cases). ogen breathing also caused significant reductions of 10.6 ± 5.4% (14.0 ± 7.62 /xm) in D, 27.3 ± 12.6% (0.48 ± 0.24 cm/sec) in V max, and 42.2 ± 10.2% (5.24 ± 2.20 /xl/min) in Q (P <.001 in all cases). ogen breathing produced significantly smaller average percentage decreases in V max and Q than O 2 breathing (P <.001 andp<.01, respectively). lthough carbogen breathing caused a smaller average decrease in D than O 2 breathing, this difference did not reach statistical significance. significant correlation was observed between the percentage decrease in D and the percentage decrease in Q during O 2 breathing (correlation coefficient, r = 0.77, P <.001, Fig. 2), whereas this correlation was not significant for carbogen inhalation (r = 0.33, P> 0.1). verage baseline values for D, V Iliax, and Q before 100% O 2 breathing were not significantly different than the average baseline values before carbogen breathing. No significant differences in average IQP, BP S, BP (1, PP, or heart rate were observed between measurements obtained immediately after the inhalation of 100% oxygen and carbogen. No significant 0 ygen ogen ygen ogen ygen ogen TBLE 2. Measurements of Venous Diameter (D), Maximum Velocity of Red Blood Cells (V max ), and Volumetric Blood Flow Rate (Q) Before and During Breathing of 100% O 2 () and ogen (), and the Decrease () in These Parameters During Breathing of the Studied Subject No. 1. ii- 2. ii- 3. ii- 4. ii- 5. ii- 6. ii- 7. ii- 8. ii- 9. ii- 10. 11. 12. ii- ii- ii- 135 119 16 175 137 38 148 122 26 132 127 5 145 17 149 124 25 158 135 23 154 26 170 137 33 18 176 151 25 147 15 D (urn) 125 11 23 147 6 124 17 152 121 31 115 13 144 15 146 139 7 168 32 165 144 21 175 13 154 8 v max (cm/sec) 1.75 1.35 0.40 1.64 0.94 0.70 1.02 0.64 1.51 1.17 0.34 1.61 1.05 0.56 1.20 0.62 0.58 1.90 1.29 0.61 2.09 1.04 1.05 1.51 0.66 0.85 2.10 1.14 0.96 2.04 0.74 1.30 1.56 0.83 0.73 1.70 1.27 0.43 1.63 1.46 0.17 1.48 1.12 0.36 1.61 1.45 0.16 2.06 1.67 0.39 1.25 0.41 1.89 1.23 0.69 2.02 1.13 0.89 1.72 1.15 0.57 1.22 0.44 1.70 0.80 0.90 1.41 1.10 0.31 Q (nl/tnin) 5.7 3.8 14.8 5.2 9.6 4.5 7.8 5.6 2.2 10.0 5.1 4.9 7.9 2.8 5.1 14.1 7.0 7.1 14.7 5.0 9.7 12.9 3.7 9.2 15.9 6.8 9.1 18.7 6.0 12.7 12.2 5.3 6.9 9.3 3.5 12.3 8.0 4.3 10.8 2.9 2.9 14.1 7.2 6.9 8.1 4.9 3.2 14.2 7.6 12.7 6.4 14.4 8.1 13.3 7.5 15.4 6.2 9.2 11.0 7.8 3.2 correlations were observed between D, V max, or Q and both blood pressure or perfusion pressure. FIGURE l. Comparison of the percentage decrease from baseline in venous diameter (D), maximum velocity of red blood cells (V inax ), and volumetric bloodflowrate (Q) during 100% O 2 breathing and carbogen breathing for each of the 12 subjects. verage percentage decreases in V max and Q were significantly smaller during carbogen breathing than during 100% O- 2 breathing (two-tailed paired Student's Mest, P <.05). DISCUSSION The results of this study show that both carbogen and 100% O 2 breathing produce significant decreases in D, V max, and Q in normal adults. Moreover, the reductions in V max and Q during carbogen breathing are significantly smaller than those observed during 100% O 2 breathing.

Effects of O 2 and CO 2 on Human Retinal Circulation 2869 ygen 10 15 20 decrease in 0 ogen 5 10 is 20 % decrease in D FIGURE 2. Percentage decrease in volumetric blood flow rate (Q) versus percentage decrease in main temporal vein diameter (D) during 100% O 2 () and carbogen breathing (B). There is a significant correlation during O., breathing (r = 0.77, P < 0.01; Q = 2 + 2.18D) and no significant correlation during carbogen breathing (r = 0.33, P> 0.10). We observed no significant difference between the vasoconstriction produced by 100% O 2 or carbogen, a finding similar to the results of Deutsch et al 10 showing a very similar vasoconstrictive effect produced by both gases. Our results show that the difference in Q reduction produced by 100% O 2 and carbogen is mainly the result of a difference in V lnax changes and less the re : suit of a difference in the changes in D measured in the main vessels. Perhaps 5% CO 2 counteracts the vasoconstriction caused by high O 2 tensions at the level of small arterioles not measured in this study. Such an effect could explain the difference in V lliax changes observed between the two gases. Because no significant differences in BP S, BP d, PP, or heart rate were detected between measurements obtained immediately after 100% O 2 and carbogen breathing, the different effects of these two gases are probably not caused by differences in their systemic effects. The significant correlation detected between the decrease in diameter of the large vessels measured and the decrease in flow during 100% O 2 breathing, and the lack of a significant correlation during carbogen breathing (Fig. 2), support the hypothesis that CO 2 may affect Q by adjusting the caliber of the smaller retinal arterioles. The average baseline values for D, V max, and Q in our study are similar to those previously reported. 14 The average decrease in Q of 56% produced by 100% O 2 inhalation is also within the range of that observed by Riva et al 8 using BLDV. The 14% decrease in D during 100% O 2 breathing is also close to values obtained by fundus photography. 15 " 5 Using the blue field simulation technique, which assesses macular circulation, Sponsel et al 11 reported decreases of 20% in leukocyte velocity with 100% O 2 breathing and an increase of 26% in leukocyte velocity during carbogen breathing. These results also suggest that retinal blood flow during carbogen breathing is greater than it is during oxygen breathing. In conclusion, the present study demonstrates that in the normal retina, the decreases in V inas and Q produced by carbogen breathing are significantly smaller than those caused by 100% O 2 breathing. These results suggest that carbogen is associated with a better perfusion of the inner retina than 100% O 2. Whether this provides a better oxygenation of the inner retina can only be ascertained by studiesjlooking at the effects of these two gases on the PO 2 6~f the inner retina. Care should be exercised when extrapolating these results, obtained in normal subjects, to what may happen in the retina of older individuals with vascular occlusion. Further studies are needed to elucidate whether carbogen may indeed provide better perfusion or better oxygenation of the inner retina, or both, than 100% O 2 in patients with retinal artery occlusion. Key Words laser Doppler velocimetry, carbogen, oxygen, volumetric blood flow rate, human retina cknowledgments The authors thank Joan Baine for her expert technical assistance and Dolly Scott for her preparation of this manuscript. References 1. ugsburger JJ, Magargal LE. Visual prognosis following treatment of acute central retinal artery obstruction. Br J Ophthalmol. 1980;64:913-917. 2. Ffytche TJ. rationalization of treatment of central retinal artery occlusion. Trans Ophthalmol Soc UK. 1974;94:468-479. 3. merican cademy of Ophthalmology. Basic and Clinical Science Course. Sec. 4. San Francisco: uthor; 1991:39. 4. Flower RW, Patz. The effect of hyperbaric oxygen on retinal ischemia'. Invest Ophthalmol Vis Sci. 197l;10:605-616. 5. lder V, BenNun J, Cringle SJ. PO 2 profiles and oxygen consumption in cat with an occluded retinal circulation. Invest Ophthalmol Vis Sci. 1990;31:1029-1034. 6. Landers MB III. Retinal oxygenation from the choroidal circulation. Trans m Ophthalmol Soc. 1978; 76: 528-556. 7. Pournaras CJ, Tsacopoulous M, Riva CE, Roth E. Diffusion of O 2 in normal and ischemic retinas of anesthetized miniature pigs in normoxia and hyperoxia. Graefe's rch Ophthalmol. 1990;228:138-142. 8. Riva CE, Grunwald JE, Sinclair SH. Laser Doppler

2870 Investigative Ophthalmology 8c Visual Science, September 1993, Vol. 34, No. 10 velocimetry study of the effect of pure oxygen breathing on retinal blood flow. Invest Ophthalmol Vis Sci. 1983;24:47-51. 9. Hickam JB, Frayser R. Studies of the retinal circulation in man: Observations on vessel diameter, arteriovenous oxygen difference, and mean circulation time. Circulation. 1966;33:302-316. 10. Deutsch T, Read JS, Ernest JT, Goldstick TK. Effect of oxygen and carbon dioxide on the retinal circulation in man. rch Ophthalmol. 1983; 101:1278-0. 11. Sponsel WE, DePaul KL, Zetlan SR. Retinal hemodynamic effects of carbon dioxide, hyperoxia, and mild hypoxia. Invest Ophthalmol Vis Sci. 1992; 33:1864-1869. 12. Grunwald JE. Effect of topical timolol on the human retinal circulation. Invest Ophthalmol Vis Sci. 1986; 27:1713-1719. 13. Damon DN, Duling BR. comparison between mean blood velocities and center-line red cell velocities as measured with a mechanical image streaking velocimeter. Microvasc Res. 1979; 17:330-332. 14. Frayser R, Hickam JB. Retinal vasculature response to breathing increased carbon dioxide and oxygen concentrations. Invest Ophthalmol. 1964; 3:427-431. 15. Sieker HO, Hickam BJ. Normal and impaired retinal vascular reactivity. Circulation. 1953; 7:79-83. 16. Grunwald JE, Riva CE, Sinclair SH, Brucker J, Petrig BL. Laser Doppler velocimetry study of retinal circulation in diabetes mellitus. rch Ophthalmol. 1986;104:991-996.