Inappropriate investigation and management of a retinal vascular occlusion

Sir,

Gupta and co-workers¹ recently presented an interesting example of segmental macular retinal infarction whose aetiology was not discoverable despite multiple invasive tests and whose management by paracentesis was complicated by submembranous prefoveal haemorrhage and Roth's spots. From the angiographic evidence available, however, it appears highly likely that their 21-year-old patient was not the victim of an isolated macular branch arterial occlusion as your correspondents had supposed. Rather this was an example of (superior) hemisphere retinal vein obstruction with secondary inner retinal infarction in the territory of a cilioretinal arteriole. Although ‘nonischaemic’ obstruction of the central retinal vein is not infrequently complicated by cilioretinal infarction, cilioretinal ischaemia secondary to hemisphere venous obstruction has been reported only once before.²

Importantly, cilioretinal infarction from hemisphere retinal vein occlusion provides an opportunity to remove any lingering doubt that may exist as to the inter-relationship between the venous and arteriolar occlusions.^{3, 4} Failure of perfusion affects only that part of the cilioretinal circulation drained by the obstructed hemisphere vein so, of the two, the venous obstruction must be the instigating occlusion.² The basis of the association between these two vascular events is said to be the lower perfusion pressure in the inner retina supplied by cilioretinal arterioles in comparison with that in the territory of the central retinal artery,⁵ but this assertion tends to hide the true picture. Significant differential effects on perfusion are only manifest when blood circulation through the inner retina is seriously challenged (eg, by marked elevation of the pressure in the central retinal vein or during ophthalmodynamometry) and they then reflect the differing patterns of branching of the central retinal and posterior ciliary arteries.^{2, 6}

This alternative diagnosis circumvents the requirement for extensive investigation in a young patient like this, not least the futile search for a source of arterial embolism. It also obviates any need to postulate an underlying retinal microvasculopathy of systemic origin to account for the ‘decompression retinopathy’.¹ Thus, hemisphere retinal vein obstruction predicts and explains the haemorrhagic consequences of a procedure (ie, paracentesis) that is ineffectual anyway in its aim of improving cilioretinal perfusion. I say this having made the same mistake myself 30 years ago.⁷ Ocular hypertension as noted in this patient may have been a factor in determining the extent of the retinal haemorrhage developing after the sudden lowering of intraocular pressure. Moreover, it was probably pathogenic in respect of the original hemisphere venous obstruction.

Unfortunately, the loculated haemorrhage that developed at the fovea as a result of the paracentesis may well have had an adverse effect on this patient's eventual visual acuity. Spontaneous improvement in vision is otherwise the rule when only a sector of the perifoveal capillary net has had its perfusion compromised as part of a combined retinal venous and cilioretinal occlusion.⁵ This highlights the importance of recognising the true nature of such occlusions, which represent a high proportion of the vascular accidents affecting the retina of young adults, have no association with serious systemic disease, and tend to resolve without intervention.^{5, 8}