Transient Visual Loss

What Questions Should Be Asked of a Patient with Transient Visual Loss?

The most important questions that need to be addressed in the assessment of the patient with transient visual loss (TVL) include the following:

Does Monocular TVL Occur Only in Certain Positions of Gaze (Gaze-Evoked TVL)?

Does the Visual Loss Occur After Prolonged Reading (Reading-Evoked TVL)?

Reading may also induce monocular TVL. Manor et al described a 49-year-old man with a 5-year history of dimming of central vision in the left eye provoked only during reading (Manor, 1996). An orbital apex tumor situated lateral to and above the optic nerve was found. This reading-evoked visual dimming may be a variant of gaze-evoked TVL. The optic nerve, displaced laterally and superiorly and stretched by the act of reading, may have been compressed between the tumor and the contracted inferior rectus muscle. Thus, orbital neuroimaging is appropriate in patients with reading-induced TVL.

Intermittent angle closure glaucoma may cause TVL, and reading-induced TVL has been reported in one case. O’Sullivan et al described a 66-year-old patient with episodes of monocular TVL lasting 3 minutes to several hours that were precipitated by reading, writing, or watching television (O’Sullivan, 1995). Ophthalmologic exam was normal but reading over 4 hours induced corneal edema, a poorly reactive semidilated pupil, and a shallow anterior chamber with intraocular pressure of 50 mm Hg. The intermittent angle closure glaucoma and the patient’s symptoms were treated successfully by iridotomies.

Do the Episodes of Monocular TVL Last Seconds?

Sharma et al found the sensitivity and specificity of visible retinal emboli for the detection of hemodynamically significant (defined as greater than or equal to 60%) carotid stenosis to be 39% and 68%, respectively, in patients with acute retinal artery occlusion (Sharma, 1998). The presence of a visible embolus generated a likelihood ratio of 1.24, whereas the absence of a visible embolus generated a likelihood of 0.88. The authors concluded that the presence of a visible embolus is a poor diagnostic test for the detection of hemodynamically significant carotid artery stenosis in the setting of acute retinal artery occlusion. Klein et al described the prevalence at baseline and the 5-year incidence of retinal emboli in the Beaver Dam Study. They reported the associated risk factors, the relationship of retinal emboli at baseline to stroke, and ischemic heart disease mortality in these patients. The study consisted of 4,926 patients, aged 43 to 86 years at baseline (Klein, 1999). The prevalence of retinal emboli at baseline was 1.3% and the 5-year incidence was 0.9%. The prevalence of retinal emboli was associated with high pulse pressure, hypertension, diabetes mellitus, past and current smoking, cardiovascular disease, and the presence of retinopathy. Patients with retinal emboli had a significantly higher risk of dying with stroke than those without retinal emboli.

TVL may also occur from ocular hypoperfusion rather than embolization. In some patients, monocular TVL may occur when the patient is exposed to bright light. These patients usually have severe, ipsilateral carotid occlusive disease. Bilateral, simultaneous TVL induced by exposure to bright light may rarely occur with bilateral severe carotid stenosis or occlusion (Kaiboriboon, 2001). The light-induced TVL probably reflects the inability of a borderline ocular circulation to sustain the increased retinal metabolic activity associated with light exposure. Alternating transient visual loss to bright light has also been described with giant cell arteritis (Galetta, 1997).

One prospective study assessed the clinical features of monocular TVL and the likelihood of atherosclerotic lesions of the internal carotid artery (ICA) (Donders, 2001). Of the 337 patients, 159 had a normal ICA on the relevant side, 33 had a stenosis of 0 to 69%, 100 had a stenosis of 70 to 99%, and 45 had an ICA occlusion. An altitudinal onset or disappearance of symptoms was associated with atherosclerotic lesions of the ipsilateral ICA. A severe (70 to 99%) stenosis was also associated with duration of TVL between 1 and 10 minutes, and with a speed of onset in seconds. An ICA occlusion was associated with attacks being provoked by light, an altitudinal character, and the occurrence of more than 10 attacks.

Postprandial transient visual loss has also been described (Levin, 1997). In one patient, episodes of splotchy visual loss occurred unilaterally on the left 1 hour after eating her largest meal of the day. The episodes lasted approximately 3 hours and were occasionally accompanied by numbness and weakness of the contralateral arm. Severe left carotid stenosis was noted. In a second patient, blotchy bilateral transient visual loss episodes lasting 2 minutes to 1.5 hours were precipitated by eating or standing from a sitting or lying position. This second patient was found to have complete occlusion of the right carotid artery and moderate stenosis of the left carotid artery. The authors proposed that postprandial visual loss may be a symptom of critical carotid stenosis, with retinal and choroidal hypoperfusion probably caused by a combination of mesenteric steal, decreased cardiac output, and abnormal vasomotor control (Levin, 1997).

Venous stasis retinopathy (hypotensive retinopathy), associated with severe carotid or ophthalmic artery occlusive disease, may also be associated with TVL (Gass, 1997). This syndrome is characterized by visual loss and ischemic retinal infarction often accompanied by signs of ciliary artery obstruction, pallor of the disc, and hypotony. Venous stasis retinopathy may simulate Purtscher’s retinopathy (multifocal areas of ischemia) and be associated with a variety of fundus pictures (Gass, 1997):

Venous stasis retinopathy may be difficult to differentiate from central retinal vein occlusion (CRVO). Helpful differentiating features include the following:

The ocular ischemic syndrome (Gass, 1997; Malhotra, 2000) is a progressive disorder due to hypoperfusion of eye that may be associated with TVL and ocular discomfort or frank pain localized to the orbit and upper face that is often decreased when the patient lies down. Rubeosis iridis in an older nondiabetic patient without evidence of venous obstructive disease or other predisposing cause is suggestive of the ocular ischemic syndrome. In persons over the age of 50 with new-onset iritis, the possibility of ocular ischemic syndrome should be considered. It is usually due to atherosclerotic carotid or ophthalmic artery disease. Other less common causes for venous stasis retinopathy and the ocular ischemic syndrome include giant cell arteritis, carotid artery dissection, cavernous sinus thrombosis, Takayasu’s disease, fibromuscular dysplasia, mucormycosis, herpes zoster ophthalmicus, myelofibrosis, vasospasm, and postaneurysm repair (Borruat, 1993; Casson, 2001; Gupta, 1990; Hamed, 1992; Hwang, 1999; Lewis, 1993; Meire, 1991; Winterkorn, 1995; Zimmerman, 1995). Four of seven patients with maxillofacial arteriovenous malformations (AVMs) that had been treated previously with proximal ligation of the supplying external carotid artery had signs of ocular ischemia (Andracchi, 2000). These four patients had significant ophthalmic artery supply by the malformations, suggesting that when the ophthalmic arterial blood supply is recruited, an ophthalmic artery “steal” phenomenon occurs, causing ocular ischemia. This “steal” may be precipitated or worsened by previous surgical proximal ligation of external carotid arterial branches that are potential collaterals with the ophthalmic artery but fail to occlude the arteriovenous shunt.