Sir,

We appreciate the opportunity to respond to the letter by Salt and Cordeiro. We agree that there are less invasive means of continuously measuring the vitreous glutamate level in an experimental model of retinal ischaemia (eg, microdialysis). As discussed in our report,1 several studies have used such technique to measure the vitreous glutamate level in various animal models of ocular ischaemia. Briefly, one study found gradual elevation of vitreous glutamate following ocular ischaemia to the peak of 6.7 times the preischaemia level in the cat.2 Another study found only transient elevation of vitreous glutamate up to seven times the preischaemia level in the rabbit.3 However, a third study failed to show any increase in the vitreous glutamate in a rabbit ocular ischaemia model.4 Another study not cited in our report monitored retinal glutamate levels continuously in real time using a dialysis electrode in a rat model of ocular ischaemia, and also failed to find glutamate elevation.5 In fact, the retinal glutamate level decreased during ischaemia, which is consistent with our results (see Table 11). Interestingly, in the same experiment, the authors induced brain ischaemia simultaneously and measured almost five-fold elevation in the brain glutamate level.5 The authors concluded that slower depletion of ATP in the retina compared to the brain allowed the retina to maintain the physiologic glutamate level and a longer tolerance to ischaemia.

How do we reconcile these disparate results? We have outlined some of the possible reasons in the report.1 One plausible explanation lies in the ability of the retinal cells’ reuptake of released glutamate through glutamate transporter. If there is sufficient reuptake into the neurons and glia during ischaemia, one may not observe elevated glutamate levels. On the other hand, if the ischaemia overwhelms the reuptake mechanism through ATP depletion and depolarization, one may see a rise in glutamate levels. In our primate model of central retinal artery occlusion, ischaemia affects only the inner retina and spares the outer retina and choroid. Such partial retinal ischaemia may allow sufficient reuptake of glutamate through intact functioning of glutamate transporter system in the retina. Indeed, there is evidence that retinal glutamate transporter activity can persist in mild ischaemic conditions in vitro.6

In our report, we limited the scope of discussion to glutamate excitotoxicity and acute retinal ischaemia. The accompanying editorial by Lotery7 and the letter by Salt and Cordeiro extend the discussion into glutamate excitotoxicity and glaucoma. Unlike the large body of literature supporting the role of glutamate excitotoxicity in acute ischaemia, the role of glutamate excitotoxicity in glaucoma was based on a handful of reports that showed elevation of vitreous glutamate levels in human and animal models of glaucoma.8, 9 Unfortunately, subsequent investigations could not corroborate the initial findings when they failed to detect elevated level of vitreous glutamate in both human and animal models of glaucoma.10, 11, 12 Thus, the original evidence that stimulated the theory of glutamate excitotoxicity in glaucoma is now in serious doubt. Salt and Cordeiro, and indeed many others in the glaucoma community, are asking whether it is still possible that glutamate excitotoxicity plays a significant role in glaucoma. The answer is unclear. What is clear is that additional, reproducible, experimental support will be required for glutamate excitotoxicity to be accepted as a significant factor in glaucoma development.