Sir,

I would like to thank Professor McLeod for the interest shown in this case report and for comments regarding the nature of arterial occlusion in eyes with central retinal vein occlusion (CRVO).

The reason that led me to conclude that this was a branch artery as opposed to a cilio-retinal artery was the fact that the affected vessel filled with fluorescein dye at 9.8 s. This was well after the choroidal flush of the angiogram when a cilio-retinal vessel would have filled with dye. This time (9.8 s) is consistent with the early arterial phase of the fluorescein angiogram.1 A closer inspection of Figure 3a also shows the proximal inferotemporal, superonasal and inferonasal arterial branches containing fluorescein simultaneously with the affected vessel. This indicates that the phase difference of the filling vessel and the other retinal arterial branches was not very significant, as would be the case in cilio-retinal and central retinal vessel filling. This difference is usually in the order of 1–3 s.1 All four major retinal arterial branches filling with dye simultaneously is more consistent with a common origin from the central retinal artery as opposed to an origin from the posterior ciliary (in the case of cilio-retinal arteries) and central retinal arteries, in which the time difference of dye filling is at least 1–3 s.1 This was not observed in this case where the time difference was not measurably more than 1 s.

The affected vessel also traces back to the bifurcation of the major retinal vessels (Figure 3a–c) at the centre of the optic disc head. This is readily visible in Figure 3 in spite of the presence of a dilated central retinal vein. Cilio-retinal arteries usually arise from the rim of the optic cup at the lateral border of the optic nerve,2,3 which was not observed in this patient.

The occluded vessel filled with fluorescein before any other branch, but this phase difference was very small, as mentioned earlier.4,5 Ordinarily, this should not be so, and in the light of the relatively short phase difference of the dye front, the difference in perfusion pressure between vessels would have to be small. Perhaps the explanation could be anomalous branching of the main central artery with this particular arterial branch arising proximally within the optic nerve head before other branches were given off. This could explain why this vessel filled with dye just before the other main branches.

Central collateral branches of the central retinal artery are known to occur. These vessels arise proximally within the optic nerve and may pass with the central artery towards the lamina cribrosa.2 It has already been pointed out that this can only be confirmed anatomically. However, there is an anomalous pattern to the arterial branching in this eye at the optic nerve head, and the optic disc in the affected eye was smaller compared to the disc on the other side. A proximal branching origin would also support the theory of vascular compression from optic disc swelling, since the said vessel would have to traverse the lamina cribrosa of the optic nerve, where compression would be likely in a swollen optic disc. Anomalous branching of retinal arteries is a rare but recognised cause of arterial occlusion in the eye.6 This was also the explanation given as to the possible aetiology in this case report, and the possibility of anomalous or collateral arterial branching should be considered.

The effect of retinal vessel autoregulation may also have contributed to the affected vessel filling slightly ahead of the other vessels. In the presence of hypoxia, as for example in an area of arterial occlusion, metabolic autoregulatory mechanisms in retinal vasculature lead to an accumulation of vasodilatory metabolites.3 As a consequence, arteriolar tone is reduced and vascular resistance reduces accordingly.3 This would result in a greater relative blood flow in this particular vessel, allowing it to fill ahead of the other retinal branches. Choroidal blood flow is not autoregulated.3

Despite thorough investigations, the exact pathophysiologic mechanism of the vascular insult in this particular case remains unknown.