Study of retinal vessel oxygen saturation in ischemic and non-ischemic branch retinal vein occlusion

Abstract

Aim: To explore how oxygen saturation in retinal blood vessels is altered in ischemic and non-ischemic branch retinal vein occlusion (BRVO).

Methods: Fifty BRVO eyes were divided into ischemic (n=26) and non-ischemic (n=24) groups, based on fundus fluorescein angiography. Healthy individuals (n=52 and n=48, respectively) were also recruited as controls for the two groups. The mean oxygen saturations of the occluded vessels and central vessels were measured by oximetry in the BRVO and control groups.

Results: In the ischemic BRVO group, the occluded arterioles oxygen saturation (SaO2-A, 106.0%±14.3%), instead of the occluded venule oxygen saturation (SaO2-V, 60.8%±9.4%), showed increases when compared with those in the same quadrant vessels (SaO2-A, 86.1%±16.5%) in the contralateral eyes (P<0.05). The oxygen saturations of the central vessels showed similar trends with those of the occluded vessels. In the non-ischemic BRVO group, the occluded and central SaO2-V and SaO2-A showed no significant changes. In both the ischemic and non-ischemic BRVOs, the central SaO2-A was significantly increased when compared to healthy individuals.

Conclusion: Obvious changes in the occluded and central SaO2-A were found in the ischemic BRVO group, indicating that disorders of oxygen metabolism in the arterioles may participate in the pathogenesis of ischemic BRVO.

Keywords: branch retinal vein occlusion; hypoxia; ischemia; oximetry; oxygen saturation.

Figures

Figure 1. The classification of BRVO based on FFA and oximetry image

A: The left pseudocolor fundus image of non-ischemic BRVO patient, shows normal retinal oxygen saturation: SaO2-A, 92.8%, SaO2-V, 56.6%; the right FFA image shows the hemorrhage, and non-perfusion area of retinal capillary ≤5 DD; B: The left pseudocolor fundus image of ischemic BRVO patient, shows increased retinal oxygen saturation: SaO2-A, 99.2% SaO2-V, 67.1%; the right FFA image shows the hemorrhage, non-perfusion area of retinal capillary >5 DD, and retinal neovascularization.

Figure 2. Oxygen saturation map and fundus image of a patient with BRVO

We performed two kinds of analysis. First kind, the vessel segments between the two circles (approximately central retina) were selected for analysis; Second kind, the first- and second- degree vessels in each four quadrant were selected for analysis. Occluded vessels were strictly chosen.

Figure 3. Comparison of oxygen saturation for occluded vessels with unaffected vessels and contralateral eye in BRVO groups

A: Significantly increased SaO2-A and SaO2-AV can be seen in occluded vessels in ischemic BRVO group (P<0.01), SaO2-V in occluded vessels a little higher than unaffected vessels in ischemic BRVO eye (though no significant difference); B: No significant differences were found between SaO2-V, SaO2-A and SaO2-AV within non-ischemic BRVO group. aStatistically significant difference P<0.05; bStatistically significant difference P<0.01.

Figure 4. The venules oxygen saturation (%) in ischemic and non-ischemic BRVO group

A: SaO2-V in occluded vessels is fluctuant (range 51.4%-71.2%), a little higher than unaffected vessels in ischemic BRVO eye (though no significant difference), also comparable to correspond quadrant in contralateral eye; B: SaO2-V in occluded vessels is rather more fluctuant (range 43%-72%) in non-ischemic BRVO eye, higher than unaffected eye, also comparable to contralateral eye. The lines connect measurements on the same patients. Measurements, which may be affected by several confounding factors, have been excluded.

Figure 5. Comparison of oxygen saturation and vessel diameter in BRVO groups with healthy individuals in central retina

Significantly increased SaO2-A and SaO2-AV can be seen in both ischemic and non-ischemic BRVO group compared to healthy individuals (P<0.05). No significant difference found between ischemic BRVO group and non-ischemic BRVO group (P>0.05). aStatistically significant difference P<0.05; bStatistically significant difference P<0.01.

Figure 6. The scatter plots and fitted curves for SaO2-V to age in BRVO group

The age was plotted on the X-axis, and the oximetry data (SaO2-V) are plotted on the Y-axis. The fitted curve for SaO2-V to age was shown in red (R2=0.06, y= -1.64x2+0.01x+37.47), indicating a negative correlation between SaO2-V and age.