Departments of Ophthalmology and of Neurology and Neuroscience, New York Presbyterian Hospital, Weill Cornell Medical College, New York, New York.
Address correspondence to Jacqueline M. S. Winterkorn, Department of Ophthalmology, New York Presbyterian Hospital, Weill Cornell Medical College, New York, NY; E-mail: email@example.com
Transient monocular visual loss remains a condition of uncertain and controversial pathophysiology. Its causes include inadequate blood flow, hypercoagulability, emboli or thrombi, and vasospasm. At various times in the medical history of the past hundred years, different hypotheses have been popular. After observing particulate matter passing through the retinal vascular bed in 1952, Fisher (1) suggested that most cases of transient monocular visual loss resulted from embolism. Recently, Trobe (2) reported that Fisher believes that the role of emboli has been overstated and that low perfusion of the retinal vascular bed is the predominant etiology. Nevertheless, the embolic theory has prevailed, and vasospasm has remained out of vogue for decades.
Although the signs of vasospasm-narrowing of the retinal vessels, boxcarring, and retinal blanching-have been observed often, vasospasm in the retina has been photographically documented in fewer than 10 patients (3-8). Vasospasm may occur as a primary event or secondary to other causes of transient monocular visual loss. It may be triggered by stress, cold, hypertension, hypotension, and low flow induced by the passage of embolic material. Winterkorn and Burde (9) published a photograph showing a narrowed artery proximal to a retinal embolus, which they attributed to vasospasm in reaction to irritation from passage of the embolus.
The notion that visual loss in the anterior visual pathway can be ascribed to migraine derives from the long-supplanted concept of migraine as a vasospastic process (10). According to the vasospastic theory of migraine, vasoconstriction explains the aura and vasodilation explains the subsequent headache. But the vasospastic paradigm cannot account for the details of blood flow during migraine (10). If migraine were a vasospastic process, its clinical manifestations would correlate in time with vasoconstriction and vasodilation of the cortical vessels, which they do not (10).
The hemodynamic changes that occur with migraine are not due to vasospasm but rather to changes in neuronal activity and platelet function (11). In migraine, oligemia occurs in multiple vascular territories in concert with a slowly spreading wave of cortical depression as described in animal models by Leão (12).
Thus, migraine affects the visual cortex and causes homonymous visual field loss. The pathophysiology of migraine would not explain monocular visual loss. Acknowledging the fact that patients often have difficulty discriminating homonymous from monocular visual loss, physicians' misapplication of the term “migraine” to a retinal mechanism confuses the issue further. In the retina, it is probably vasospasm, not migraine that causes unexplained transient monocular visual loss, especially in otherwise healthy young individuals (9).
The term “retinal migraine” is anatomically inconsistent, referring to a cortical process as though it were occurring in the retina. Unfortunately, the International Headache Society (IHS) has defined “retinal migraine” as two attacks of fully reversible monocular visual loss associated with migraine headache. Grosberg et al (13) recently extended the definition to include patients with persistent monocular visual loss, hypothesizing that a migrainous process can cause persistent retinal damage-“an ocular form of migrainous infarction.” Of course, migraine is so prevalent that it could occur in a patient who also has retinal vasospasm or retinal vascular occlusion. However, migraine is a different process occurring in a different location. The basis for migraine is neuronal instability rather than vascular instability. Indeed, the prevailing view articulated by Goadsby (11) is that migraine arises from changes in brainstem regions involved in sensory modulation. Patients with vasospastic conditions such as Raynaud digital symptom or Prinzmetal coronary angina may be susceptible to development of ocular vasospasm (6,9), and this suggests an underlying pathophysiology for vasospastic conditions. If only because migraine so rarely occurs together with these conditions, we should be looking for a different underlying cause for vasospasm.
In this issue of the Journal, Hill et al (14) apply strict IHS criteria in a broad-based review of the reported cases of transient monocular visual loss, finding that retinal migraine is, at best, exceedingly rare. Only 5 of 142 cases they reviewed qualify as definite retinal migraine under the IHS criteria. The authors attribute the other cases of transient monocular visual loss to retinal vasospasm.
Hill et al (14) have done a good job of helping to rehabilitate vasospasm as a cause of monocular visual loss. They argue that such visual loss is most often not accompanied by migrainous headache or aura. They point out that although cortical spreading depression is an attractive hypothesis for the pathophysiology of migraine, no clinical correlation has been made between retinal spreading depression and monocular visual loss. As the authors also remind us, spreading depression has been observed in vitro in the avascular retinas of frogs and chicks but has never been seen in vascularized mammalian retinas (14).
These are good arguments, but it is important to keep in mind the fact that retinal vasospasm remains a diagnosis of exclusion except in the very rare patient in whom it is caught in the act by ophthalmoscopy. It is especially important not to ascribe persistent visual loss to migraine but to pursue the cause, with migraine as one of the clues. A migraine diathesis may be a marker for an underlying defect such as hypertension, hypercoagulability, lupus erythematosus, or mitral valve prolapse and therefore might occasionally co-occur with vascular occlusion or retinal vasospasm. In diagnosing vasospasm, care should be taken to address these conditions, as well as optic disc drusen, smoking, and glaucoma. When vasospasm remains after other conditions are excluded, confirmatory treatment with calcium channel blockers may be indicated (6,15).
“Retinal migraine,” with its cousins “ophthalmic migraine” and “ocular migraine,” is an oxymoron. It is misleading and should not be part of the medical vocabulary.
1. Fisher CM. Transient monocular blindness associated with hemiplegia. Arch Ophthalmol
2. Trobe JD, Fisher CM. The master of clinicopathologic correlation. J Neuroophthalmol
3. Humphrey WT. Central retinal artery spasm. Ann Ophthalmol
4. Kline LB, Kelly CL. Ocular migraine in a patient with cluster headaches. Headache
5. Burger SK, Saul RF, Selhorst JB, et al. Transient monocular blindness caused by vasospasm. N Engl J Med
6. Winterkorn JMS, Kupersmith MJ, Wirtschafter JD, et al. Treatment of vasospastic amaurosis fugax with calcium channel blockers. N Engl J Med
7. Bernard GA, Bennett JL. Vasospastic amaurosis fugax. Arch Ophthalmol
8. Petzold A, Islam N, Plant GT. Video reconstruction of vasospastic transient monocular blindness. N Engl J Med
9. Winterkorn JMS, Burde RM. Vasospasm-not migraine-in the anterior visual pathway. Ophthalmol Clin N Am
10. Lauritzen M, Olesen J. Regional cerebral blood flow during migraine attacks by xenon 133 inhalation and emission tomography. Brain
11. Goadsby PJ. Recent advances in understanding migraine mechanisms, molecules and therapeutics. Trends Mol Med
2006(Nov 30);Epub ahead of print.
12. Leão AAP. Spreading depression of activity in the cerebral cortex. J Neurophysiol
13. Grosberg BM, Solomon S, Friedman DI, et al. Retinal migraine reappraised. Cephalalgia
14. Hill DL, Daroff RB, Ducros, A, et al. Most cases labeled as “retinal migraine” are not migraine. J Neuroophthalmol
15. Winterkorn JMS, Teman AJ. Recurrent attacks of amaurosis fugax treated with calcium channel blocker. Ann Neurol