Search

In miniature pigs, retinal veins were experimentally occluded using argon laser coagulation. Microvascular modifications leading to retinal hemorrhages and retinal edema were observed some hours after the occlusion. These lesions resolved progressively within 3 weeks after the occlusion, but in most cases ischemic retinal territories persisted. Preretinal partial pressure of oxygen (PO2) measurements, using double barrelled O2-sensitive microelectrodes, showed that all the ischemic areas were indeed hypoxic. In half of the experiments, preretinal and intravitreal new vessels grew on the ischemic territories. Tissue hypoxia appears to be a key step in triggering neovascularization. However, the critical level of hypoxia was not determined.

Using laser Doppler flowmetry, slow variations in velocity, volume, and flux of red blood cells in the optic nerve head (ONH), choroid, and retina of the anesthetized minipig have been demonstrated. The variations of velocity and volume were highly regular and vigorous in the ONH and had frequencies ranging from 2.5 to 4.5 cycles/min. The flux variations were smaller or absent due to a phase shift of approximately 180 degrees between the volume and velocity changes. The volume fluctuations were synchronized to those of the PO2 measured in the vitreous, at approximately 50 microns from the surface of the ONH tissue. The fluctuations were less regular in the choroidal and the retinal vessels and their frequencies were higher than those in the ONH. The lack of correlation between the fluctuations in the ONH and those in the retinal and choroidal vessels points toward a local mechanism. The changes in blood volume in the ONH, the phase shift between volume and velocity changes, and the predominance of venules at the surface of the ONH are three factors suggesting that this mechanism involves a change in the diameter of the venules rather than in the arterioles.

Transretinal pO2 profiles were recorded during normoxia and hyperoxia in normal and ischemic retinal regions in anesthetized miniature pigs, using double-barrelled recess-type microelectrodes. In normoxia and hyperoxia, pO2 in the normal region decreased from the inner retina and the choroid toward the midretina, indicating that the choroid cannot supply O2 to the entire normal retina. Preretinal and transretinal pO2 measurements in ischemic regions following laser occlusion of a retinal branch vein showed that in normoxia the direction of pO2 gradients prevents O2 diffusing from the choroid to reach the inner retina. This explains why the ischemic regions remain hypoxic. On the contrary, during hyperoxia the intraretinal pO2 gradient indicates an O2 flux from the choroid to the inner retina, resulting in a marked increase in preretinal pO2 in the affected regions. Hence hyperoxia could be a useful tool for restoring the oxygen supply to the inner, hypoxic retinal layers; unfortunately it cannot be used in clinical practice. Parabulbar injections of dexamethasone induce a transitory increase in preretinal pO2, probably by reducing the outer retinal consumption of O2 and thus allowing O2 which diffuses through the choroid to reach and restore the inner retinal hypoxia; clinical experience has shown that parabulbar dexamethasone injections may be effective in the treatment of venous branch occlusion.

When measured with double-barreled ion-selected microelectrodes, delta K+0 is found to be heterogeneously distributed over the retinal surface. A transient fall (less than or equal to 200 microM) in delta K+0 near an arteriole wall and a transient increase (less than or equal to 500 microM) in zones free of arterioles were observed during light stimulation. In view of these findings the authors assume that the former is due to K+ uptake and transport by perivascular astrocytes, and that the latter is due to K+ siphoning by Müller cells.

Transretinal PO2 profiles were recorded during normoxia and hyperoxia in normal and ischemic retinal territories in anesthetized miniature pigs using double barrelled recess type microelectrodes. In normoxia and hyperoxia the PO2 in the normal territory decreased from the inner retina and the choroid towards the mid-retina, indicating that the choroid cannot supply O2 to the whole normal retina. Preretinal and transretinal PO2 measurements in ischemic territories following a laser occlusion of a retinal branch vein demonstrated that in normoxia the direction of PO2 gradients prevents O2 diffusing from the choroid to reach the inner retina. This explains why the ischemic territories are hypoxic. In the contrary, during hyperoxia the intraretinal PO2 gradient indicates an O2 flux from the choroid to the inner retina resulting to marked preretinal PO2 increase at the affected territories. We proposed the hypothesis that in the ischemic retinas the hyperoxia does not induce a rise of the O2 consumption of the outer retina. Hence hyperoxia could be a useful tool to restore the oxygenation of the inner hypoxic retinal layers.

Transretinal PO2 profiles were recorded with O2-sensitive microelectrodes in the normal retina and in ischemic retinal foci induced by the occlusion of a retinal branch vein with argon laser photocoagulation in anesthetized miniature pigs. In the normal retina there are two PO2 gradients: one from the inner retina and the other from the choroid, both directed toward the middle of the retina. Both PO2 gradients persisted during hyperoxia. Thus, even in hyperoxia, the choroid does not supply the whole thickness of the normal retina with O2. Preretinal and transretinal PO2 measurements in ischemic inner retinal foci showed the existence of two PO2 gradients in steady-state systemic normoxia, as did those in the normal retina. This finding indicates that even in ischemia the choroid does not supply O2 to the inner retina; as a result, tissue hypoxia is maintained. During systemic hyperoxia, the intraretinal PO2 measurements in the ischemic foci showed only one gradient going from the choroid toward the inner retina. This gradient indicates that under these conditions, the choroid can supply O2 to the entire thickness of the ischemic retina. Extending a previously formulated hypothesis, we propose that in the ischemic retina as opposed to the normal retina, hyperoxia does not induce an increase in the O2 consumption of the outer retina. This suggestion could explain the rise in PO2 in the inner ischemic retina during hyperoxia.

Hypercapnia causes vasodilatation of retinal arterioles. Prostaglandin E1, injected close to retinal arterioles from the vitreal side cause vasodilatation apparently similar to that caused by hypercapnia. An inhibitor of prostaglandin synthase (indomethacin) was injected into the ocular circulation. There was a reversible inhibition of the retinal vasodilatation normally induced by hypercapnia. Indomethacin injected close to retinal arterioles from the vitreal side causes reversible vasoconstriction and inhibits the vasodilatory effect of hypercapnia. It is concluded that prostaglandin E1 satisfies three criteria for a candidate for a mediator of hypercapnia-induced arteriolar dilatation.

Experimental occlusion of a retinal vein leads to retinopathy in the affected drainage area. In some cases the formation of ischemic regions, where the retina is completely destroyed, is typical for the development of this retinopathy. Laser photocoagulation restores the physiological preretinal oxygen pressure in these hypoxic regions. It is possible that the rise in O2 pressure which has been observed is caused by destruction of pigment epithelium and the photoreceptor layer, as oxygen diffuses freely from the choroid to the inner retina without being consumed by the outer layers.

Laser Doppler velocimetry was performed on retinal vessels of a cat using a linearly polarized He-Ne laser as the incident beam. The diameter of the vessels measured was <120 microm. Measurements show that if double transmission of the laser light through a given retinal vessel can be prevented, the Doppler shift power spectra have the shape theoretically expected from calculations based on a single scattering model and parabolic velocity profile of the red blood cells.

Treatment of a retinal vascular occlusion consists in restoration of normal oxygenation to the affected area. An experimentally induced increase in arterial CO2 provokes an increase in preretinal CO2 in a hypoxic retinal area which has developed after venous obstruction, but it is ineffective in anoxic areas of an arterial occlusion. However, clinical application of this treatment offers no benefit for the patient. Hyperoxia caused by respiration of a gas mixture of 5% CO2 and 95% O2 is also ineffective. On the other hand, the results of recent studies in which blood is replaced with fluorocarbon solutions are promising. Finally, it is pointed out that in cases of venous occlusion laser coagulation leads to reoxygenation of the inner hypoxic retina.

Aging exudative macular degeneration frequently results in a fibrovascular scar which entails the loss of central vision. This scar implies an often large and deep scotoma which makes it very difficult for the patient to be fitted with low-vision aids. A Goldmann static perimetry study of retinal sensitivity in the treated areas showed that krypton laser at liminar doses does not induce deep scotomas.

A constant oxygen supply from the retinal and choroidal vascular system is essential for the maintenance of the retinal electrical activity. Arterial or venous retinal obstruction modifies the hemodynamics in the retinal circulation and produces hypoxia in the affected area. In the case of arterial obstruction, severe hypoxia (anoxia) rapidly (in less than 45 min) produces irreversible cell damage in the inner area of the retina. With venous obstruction, the hypoxia is moderate and the retinal damage is partial. In some cases, hypoxia regresses because the formation of collateral vessels most probably provides adequate perfusion. In cases of irreversible venous obstruction, insufficient formation of collateral vessels results in the development of ischemic regions in the retina and severe hypoxia. In these cases preretinal neovascularization may develop.

Intraretinal oxygen tension (pO2) and local electroretinogram (ERG) were simultaneously measured in miniature pigs using double-barreled recess type microelectrodes. Transretinal pO2 profiles were recorded during normoxia and hyperoxia in areas close to (juxta-arteriolar) or far from (intervascular) retinal arterioles. In normoxia, in both areas, the pO2 decreased from the inner retina and the choroid towards the middle of the retina. In the inner retina the gradient of the juxta-arteriolar pO2 profile was steeper than that of the intervascular profile. This characteristic persisted during the breathing of 100% O2. Analysis of the pO2 profiles shows that, even in hyperoxia, the choroid cannot supply O2 to the whole retina. The results also support the conclusions of previous work (Riva, Pournaras and Tsacopoulos, 1986) indicating that in the normal retina it is not the O2 diffusing from the choroid to the retinal arterioles the induces vasoconstriction of these vessels. In the miniature pig this constriction appears to maintain inner retina tissue pO2 at a constant level during hyperoxia. From the pO2 transretinal profiles and previously published choroidal O2 fluxes and pO2 values near retinal vessels an explanatory working hypothesis is formulated according to which O2 consumption (qO2) of the outer retina increases during hyperoxia in the miniature pig.

Choroidal neovascularization is considered to be the cause of exudate or subretinal hemorrhage formation in the evolution of different types of macular degeneration (exudative senile macular degeneration, idiopathic macular degeneration, presumed ocular histoplasmosis, angioid streaks, etc. The only curative treatment which has proved to be successful is the destruction of subretinal neovascularisation by photocoagulation. The effect of laser photocoagulation depends on the radiation wavelength emitted and the absorption spectrum of the target tissue. It is for this reason that the red krypton laser is recommended for the destruction of subretinal neovascularizations, its action being restricted to the choroid without inflicting damage on the nerve fiber layer.