Sir,

We are grateful to Dr McLeod1 for the interest that he has shown in our paper.2 Rather, then there being a conspiracy to increase the confusion within central retinal artery occlusion (CRAO), our study demonstrates that the disease is incompletely understood. In particular, we disagree with Dr McLeod that the retinal penumbra lasts for more than 24 h. Retinal and cerebrovascular ischaemia share a number of common pathophysiological features and is reflected by the recent extended definition of stroke involving retinal as well as cerebral ischaemia.3

In vitro experiments show that when neuronal cells are deprived of oxygen, that within 5 s, there is evidence of neuronal dysfunction. Within 10 s, cell death occurs. The elegant experiment of Astrup et al4 demonstrated that the penumbra is a function of collateral perfusion and that if there was no resolution of the occlusion, then eventually the penumbra would fail, and infarction would be permanent.

In cerebral stroke using perfusion imaging, various groups have demonstrated that the ischaemic penumbra may persist beyond 24 h. However, randomised controlled trials of reperfusion therapy in acute stroke have demonstrated on that in the majority of individuals, the ischaemic penumbra only extends out to 4.5 h and at maximum, 6 h.5, 6

The misperceptions that the retinal penumbra persists for 24 h, initially backed up by observational data, led to the design of two randomised controlled trials that recruited individual with central retinal artery occlusion of beyond 6 h. The EAGLE study recruited subjects up to 19 h of symptom onset,7 while our group conducted a randomised controlled trial of intravenous tPA given to individuals within 24 h of symptom onset.8 Both of these randomised controlled trials were negative studies, however, a signal was seen in individuals who receive tPA within 6 h of symptom onset.8 This mirrors the retinal tolerance time demonstrated in Hayreh’s animal experiments of 4 h.9

Conversely, if the thrombus that initially caused a retinal artery occlusion resolves such that retinal perfusion is restored, then the symptoms will be invariably improved leading to a transient central retinal artery occlusion.

Nevertheless, we agree with Dr McLeod that the classical clinical features of central retinal artery occlusion remain more or less the same, but that various findings may give a clue as to the time of the onset or alternatively the presence of partial occlusion. These are important points given that for randomised controlled trials on the treatment of CRAO are to succeed, then individuals need to be recruited within the shortest time possible.