The accumulated oxygen deficit as an indicator of the ischemic retinal insult

We here attempt to improve quantification of the ischemic retinal insult, that is, what is imposed on the retinal tissue by ischemia, especially in experimental models of ischemia. The ischemic retinal insult initiates the ischemic retinal injury (or outcome). Accordingly, it is reasonable to assume that the better the quantification of the insult, the better the correlation with, and thereby estimation of, the injury. The insult seldom has been quantified in terms of the relevant physiological factors, especially in connection with the rate of oxygen delivery (DO₂). We here propose the accumulated oxygen deficit (AO₂D) as an indicator of the ischemic retinal insult. We hypothesized that AO₂D is correlated with the rate of oxygen metabolism measured 1 h after reperfusion following an episode of ischemia (MO₂_1_Hr). Previously, we showed that MO₂_1_Hr is related to the electroretinogram amplitude and the retinal thickness when they are measured seven days after reperfusion. We studied 27 rats, as well as 26 rats from our published data on retinal ischemia in which we had measurements of DO₂ and duration of ischemia (T) of various levels and durations. We also measured DO₂ in 29 rats treated with sham surgery. Ischemia was induced by either ipsilateral or bilateral common carotid artery occlusion or by ophthalmic artery occlusion, which gave a wide range of DO₂. DO₂ and MO₂_1_Hr were evaluated based on three types of images: 1) red-free images to measure vessel diameters, 2) fluorescence images to estimate blood velocities by the displacement of intravascular fluorescent microspheres over time, and 3) phosphorescence images to quantify vascular oxygen tension from the phosphorescence lifetime of an intravascular oxygen sensitive phosphor. Loss of oxygen delivery (DO₂L) was calculated as the difference between DO₂ under normal/sham condition and DO₂ during ischemia. AO₂D, a volume of oxygen, was calculated as the product DO₂L and T. Including all data, the linear relationship between AO₂D and MO₂_1_Hr was significant (R² = 0.261, P = 0.0003). Limiting data to that in which T or DO₂L was maximal also yielded significant relationships, and revealed that DO₂L at a long duration of ischemia contributed disproportionately more than T to MO₂_1_Hr. We discuss the potential of AO₂D for quantifying the ischemic retinal insult, predicting the ischemic retinal injury and evaluating the likelihood of infarction.