Sir,

Ozdek et al1 report details of two patients with hyperhomocysteinaemia who presented with nonischaemic occlusion of the central retinal vein (CRV) and putative concurrent occlusion of a branch retinal arteriole. They make reference to an earlier description of seven eyes similarly affected by CRV occlusion and segmental inner retinal infarction2 in which cilioretinal infarction, secondary to the CRV occlusion,3, 4 was discounted as the mechanism of ischaemia. However, a single vascular luminal obstruction is more likely than two simultaneous (but separate) occlusions, especially in young patients. The possibility of cilioretinal infarction should be seriously entertained, therefore, even if this implies that up to half of the retinal circulation must perforce derive from posterior ciliary branches of the ophthalmic artery. Indeed, in Case 1 of this latest report,1 the inferotemporal vessel supplying the territory of the infarct appears to have a ‘hook’ characteristic of a cilioretinal arteriole as it emerges from the disc rim.

In CRV occlusions, the distinction between a (simultaneous) branch arteriolar occlusion and a (consequential) cilioretinal arteriolar occlusion is not merely of academic interest. Recognition of the dependent association will spare the patient from unnecessary investigations, including the search for a source of arteriolar embolism. Moreover, therapeutic lowering of the intraocular pressure (as attempted in Case 1) runs the risk of precipitating retinal haemorrhage by increasing transmural hydrostatic pressure gradients, without any prospect of improving inner retinal perfusion.5