Anterior Ischemic Optic Neuropathy in Eyes With Optic Disc Drusen

Background  There have been anecdotal reports of anterior ischemic optic neuropathy(AION) occurring in eyes with optic disc drusen (ODD), but the clinical featuresof this condition have not been well characterized.

Objectives  To better describe the clinical features of AION associated with ODDand to compare the clinical features of this condition with those of "gardenvariety" nonarteritic AION.

Methods  We reviewed the medical records of 20 patients who experienced an episodeof AION in an eye with ODD. In 4 patients, both eyes were affected; thus,24 eyes were studied. The diagnosis of ODD was made by ophthalmoscopic identification,orbital ultrasonography, or computed tomographic scanning. We recorded age,sex, vascular risk factors, symptoms, visual acuity, visual fields, and resultsof the follow-up examination in all patients. These findings were comparedwith data from previously reported series of patients with nonarteritic AION.

Results  Our 20 patients included 14 men and 6 women (age range, 18-69 years;mean, 49.4 years). Vascular risk factors were identified in 10 patients (50%).Three patients reported episodes of transient visual loss before their fixeddeficit. The visual acuity at the initial examination was 20/60 or betterin 15 (62%) of the 24 eyes; 8 had a visual acuity of 20/20. The predominantpattern of visual field loss was an altitudinal or arcuate defect in 19 (79%)and a centrocecal scotoma in 5 (21%) of the 24 eyes. There was subjectiveworsening of vision before the initial neuro-ophthalmic examination in 11eyes (46%) and objective documentation of progression in 7 eyes (29%). Thefinal visual acuity was 20/40 or better in 13 (62%) of 21 eyes and 20/200or worse in 3 (14%) of 21 eyes.

Conclusions  Our patients were strikingly similar to those with nonarteritic AIONunassociated with drusen in regard to prevalence of vascular risk factors,pattern of visual field loss, and occurrence of a subsequent similar eventin the fellow eye. In contrast, however, patients with ODD-AION were youngerthan those with nonarteritic AION, were more likely to report preceding episodesof transient visual obscuration, and enjoyed a more favorable visual outcome.

Optic disc drusen (ODD) are a relatively common congenital disc anomalyfound in up to 2% of the population.^1,2 Theymay appear as refractile bodies on the disc surface or may be buried and producedisc elevation. When buried, the main clinical significance of drusen is theirresemblance to true papilledema. Impairment of visual acuity due to drusenis rare. In most cases, the visual deficit takes the form of slowly progressivefield loss that may resemble glaucomatous visual loss and is often unnoticedby the patient.^3,4 Occasionally,a more rapid decline of vision occurs, sometimes associated with decreasedcentral vision.^5,6 Some of theseevents are due to hemorrhage that may involve the disc, the peripapillaryretina, or the vitreous.^7,8 Inother cases, visual loss is a consequence of vascular insufficiency affectingeither the retina^9,10 or the opticnerve.¹¹

Ischemia of the optic nerve in patients with ODD may or may not be associatedwith disc edema. In patients without acute disc changes, the presumed mechanismof visual loss is infarction of the retrolaminar optic nerve.^5,12,13 Patientswith acute disc edema are assumed to have anterior ischemic optic neuropathy(AION). The association of AION and ODD has been reported, but the clinicalfeatures of this condition are not well described.^14-16

We reviewed the medical records of patients with ODD and acute visualloss examined in the neuro-ophthalmology clinics of the Midwest Eye Instituteor Indiana University Medical Center, Indianapolis, between 1983 and 2001.All patients were examined by at least one of us (V.P., R.K., A.K., or R.Y.).All patients gave written consent for the use of their clinical material forpublication. A diagnosis of AION was made in patients with acute monocularvisual loss with disc-related field loss in the involved eye and disc edema.Patients were excluded if they were not seen within 4 weeks from the onsetof visual loss, if disc edema was not observed, or if they had clinical orlaboratory evidence of systemic vasculitis. Other causes of optic neuropathy,such as compressive lesions and inflammatory or demyelinating disease, wereexcluded by medical history, physical examination, and appropriate ancillarytesting, as indicated. In all patients, the results of a complete blood countand the erythrocyte sedimentation rate were obtained. The diagnosis of ODDwas made by ophthalmoscopic identification of drusen, orbital ultrasonography,or computed tomographic scanning.

Twenty patients with acute AION and ODD were identified. In 4 patients,both eyes were affected; thus, 24 eyes were studied. We recorded age, sex,vascular risk factors (hypertension, diabetes mellitus, hypercholesterolemia,and tobacco use), other systemic disorders, symptoms, visual acuity, and resultsof visual field testing. The results of follow-up examinations were recordedwhen available. All patients underwent formal visual field testing using eitherGoldmann perimetry or an analyzer (Humphrey Automated Field Analyzer 30-2).Visual field defects were classified by their predominant pattern.

Progression of visual loss was classified as subjective or objective.Subjective progression was determined at the initial neuro-ophthalmic evaluationby direct inquiry. Objective progression was assessed by serial examinationsand defined as a decrease of 1 line or more in Snellen acuity within 1 monthof the initial examination.

Our 20 patients included 14 men and 6 women. Their ages ranged from18 to 69 years (mean, 49.4 years). Of the 24 affected eyes, 13 were rightand 11 were left. Vascular risk factors were identified in 10 patients (50%).Four patients had hypertension, 4 had diabetes mellitus, 1 had hypercholesterolemia,and 1 used tobacco. Other systemic conditions included melanoma, migraine,psoriasis, and pregnancy (1 patient with each condition). Other ocular disordersincluded chronic open-angle glaucoma in 2 patients, ocular hypertension in1, previous iritis in 1, prior cataract extraction in 1, and congenital colorblindness in 2. Visual loss was associated with mild eye pain in 5 eyes, describedas an ache, pressure, sensation, or irritation. No patient had severe painor pain with eye movement. In 4 eyes, visual loss was accompanied by a mildheadache. Three patients reported transient obscurations of vision beforeonset of the fixed deficit. All patients were examined within 4 weeks of onset(mean, 8.5 days; median, 4.5 days).

Of 24 affected eyes, the visual acuity at the initial examination was20/60 or better in 15 (62%); 8 had a visual acuity of 20/20. In 7 eyes, thevisual acuity was worse than 20/200. Visual field loss was predominantly altitudinalor arcuate in 19 (79%) of the 24 eyes. Five eyes (21%) exhibited a centrocecalscotoma.

Patients described subjective worsening of vision before the initialexamination in 11 eyes (46%) over 2 to 7 days. In 7 eyes (29%), serial examinationsshowed progression of visual loss within the first month following onset.A total of 14 eyes (58%) showed subjective or objective progression or both.

Follow-up information was available for 21 eyes, with an average intervalbetween the initial and final examination of 2.9 years (range, 6 months to16 years). Three eyes were not included in the assessment of visual outcome.In 2 eyes (1 patient), the patient was not available for reexamination. In1 eye, optic neurotomy was performed and, thus, the outcome was not consideredto represent the natural history of the disorder. At last follow-up, 13 (62%)of 21 eyes showed a visual acuity of 20/40 or better. Of the 21 eyes, 3 (14%)had a visual acuity of 20/200 or worse. Compared with the initial examination,14 of 21 eyes showed no change or an improvement of 1 line or more. Seveneyes showed worsening by 1 or more lines. Of the 6 eyes that improved, theincrease in Snellen acuity was 1 line in 1, 2 lines in 1, 3 lines in 1, 4lines in 2, and 6 lines in 1. Of eyes that worsened, the change measured 1line in 2, 2 lines in 3, 3 lines in 1, and 6 lines in 1. Of the 7 eyes thatdemonstrated initial progression of visual loss, 5 showed subsequent improvementof 1 to 7 lines.

The diagnosis of ODD was established by fundus examination in 7 eyes,by orbital ultrasonography in 15 eyes, and by computed tomography in 2 eyes.Ultrasonography and/or computed tomography were pursued in some patients becausethe disc appearance was suggestive of buried drusen (bumpy surface, scallopedmargin, or deep peripapillary hemorrhage), in others because of the absenceof vascular risk factors, in 2 because of unusually prolonged progressionof visual loss, and in 1 because of recurrence of AION in a previously involvedeye. Optic disc drusen were bilateral in 16 (80%) of the 20 patients. In 4of these patients, an episode of AION was documented in each eye. In 2 patients,the interval between events was 3 months; and in 2 others, 6 months. In nopatient were the 2 events simultaneous. In 2 other patients with bilateralODD, the clinical findings were suggestive of a previous AION episode in thefellow eye. One patient experienced a second episode of AION in an involvedeye.

A 44-year-old man experienced abrupt onset of spotty vision in the lefteye, which was present on awakening. There was no associated head or eye pain,no systemic symptoms, and no history of focal neurologic deficit. His visionwas subjectively worse the next day and then stable the day after. He wasgenerally healthy and taking no medications.

A neuro-ophthalmic examination 3 days after onset of visual loss revealeda visual acuity of 20/20 OD and 20/300 OS. He identified all 15 pseudoisochromaticplates in each eye. There was a 1.8–log unit relative afferent pupillarydefect in the left eye. Goldmann perimetry in the right eye showed slightinferonasal constriction peripherally. In the left eye, there was a superiorarcuate scotoma (Figure 1). A fundusexamination showed marked elevation of the right optic disc without opacificationof the nerve fiber layer and with no physiologic cup. The left disc was alsoquite elevated, had a bumpy surface, and showed, in addition, capillary hyperemiaand nerve fiber layer hemorrhages (Figure2). A computed tomographic scan revealed calcification within theoptic discs bilaterally, and was otherwise normal (Figure 3). The results of laboratory testing, including a completeblood count and the erythrocyte sedimentation rate, were normal.

On reexamination 3 weeks later, the patient reported no subjective changein vision but, in fact, visual acuity had improved to 20/25 OS, perhaps inpart because the patient became more adept at extrafoveal fixation. He identified10 of 17 pseudoisochromatic plates in the left eye, and the left relativeafferent pupillary defect had decreased to 0.9 log units. Goldmann perimetryshowed the same pattern of loss in the left eye, but with less central depression,including return of a small I-1-e isopter above fixation. The left disc wasless edematous. All findings in the right eye were unchanged.

Two months later, the patient experienced sudden onset of painless visualloss in the right eye, which progressed during the first 48 hours and thenstabilized. An ophthalmic examination 5 days after onset showed a visual acuityof 20/400 OD and 20/25 OS. Goldmann perimetry in the right eye showed a broadcentrocecal scotoma (Figure 4).The right disc was swollen with peripapillary nerve fiber layer hemorrhages;and the left disc was pale, with drusen visible nasally (Figure 5). The results of additional testing for coagulopathy andsystemic inflammatory disease were unrevealing.

This 59-year-old man experienced onset of painless visual loss in theleft eye associated with mild photosensitivity. There was no history of headacheor focal neurologic deficits. His ocular history was positive for bilateralODD. His medical history included obesity and borderline hypercholesterolemia.

A neuro-ophthalmic examination 2 weeks after the onset of visual lossdemonstrated a visual acuity of 20/20 OU, with normal color vision and a 0.9–logunit relative afferent pupillary defect in the left eye. Goldmann perimetryin the right eye showed enlargement of the physiologic blind spot and mildinferior depression. In the left eye, there was dense peripheral inferotemporalloss and a superior nasal scotoma (Figure6). The fundus examination revealed marked bilateral ODD with peripapillaryatrophy and anomalous branching of retinal arteries. In addition, in the lefteye, there was mild thickening of the nerve fiber layer nasally with a fewperipapillary hemorrhages (Figure 7).A fluorescein angiogram confirmed late staining of the left disc but not theright. The results of a complete blood cell count and the erythrocyte sedimentationrate were normal. Reexamination 2 months later showed mild worsening of thevisual field in the left eye, with resolution of disc hyperemia and hemorrhages.Although there were no further episodes of acute visual changes, a reexamination7 years later revealed further field loss nasally and superiorly in the lefteye.

Nonarteritic AION (NAION) is a common cause of acute visual loss. Themechanism in most patients is thought to involve a decrease in perfusion pressurewithin the optic nerve head that, in susceptible individuals, leads to infarction.¹⁷ This susceptibility may be related to abnormalitiesin the vasculature and the structure of the optic disc. The particular aspectof optic disc structure that has been implicated is a small cup-disc ratio,the so-called disc at risk.¹⁸ Optic disc drusenrepresent another form of disc anomaly in which the nerve head is crowdedand cupless. Anecdotal reports^{5,14-16,19,20} ofNAION occurring in eyes with ODD suggest an association between these 2 conditionsas well. While a possible causative role has been proposed,¹⁴ otherresearchers²¹ have suggested that the coexistenceof these 2 disorders is coincidental.

In this study, we analyzed the clinical features of 20 patients withdrusen-associated AION (DAION). We hoped to provide a more complete descriptionof this disorder and to clarify its relationship to the better-defined entityof "garden variety" NAION. Our 20 patients with DAION were a mean age of 49.4years (range, 18-69 years). This contrasts with previously described populationsof patients with NAION in whom the mean age was 63.4 years.^22,23 Vascularrisk factors were identified in 50% of our patients, compared with 60% ofthe NAION patients in the Ischemic Optic Neuropathy Decompression Trial.²⁴ In addition to the primary symptom of acute monocularvisual loss, 3 patients (12.5%) reported preceding transient obscurationsof vision and 5 (21%) described mild ipsilateral eye pain. By comparison,5.4%²² and 10%^24-26 ofpatients with NAION reported these symptoms, respectively.

Four patients experienced bilateral sequential episodes of DAION, resultingin a total of 24 eyes studied. The visual acuity at the initial examinationwas 20/60 or better in 62% of the eyes (vs 31%-52% with this visual acuityin a combined NAION series²⁷). A visual acuityof 20/200 or worse was found in 29% (vs 35%-54% in the NAION series²⁷). The predominant pattern of visual field loss consistedof an altitudinal or arcuate defect in 79% and a centrocecal scotoma in 21%.In comparison, Repka et al²³ found these patternsof visual field loss in 79% and 20%, respectively; Boghen and Glaser²² found altitudinal or arcuate defects in 71%, andRizzo and Lessell²⁵ reported a centrocecalscotoma in 26%.

Patients described subjective worsening of vision before the initialneuro-ophthalmic examination in 46% of episodes. In 29% of events, furthervisual loss was documented after the initial neuro-ophthalmic examination.These percentages are comparable to those reported in the Ischemic Optic NeuropathyDecompression Trial,²⁴ in which 45% and 29%of eyes determined such subjective and objective progression, respectively.The visual acuity at the final examination was 20/40 or better in 62% of theeyes. This is in contrast to combined series²⁷ ofpatients with NAION in whom this level of final visual acuity was found in21% to 53%. A visual acuity worse than 20/200 was found in only 14% in ourseries (vs 31%-41% in the combined series²⁷).

Of 20 patients, 4 (20%) experienced bilateral sequential episodes ofDAION. The incidence of bilateral NAION has varied widely in different series²⁸ because of the nonhomogeneous patient populationand differences in the duration of follow-up. The previously reported rangeof bilaterality is from 10.5% to 73%,²⁹ witha recent estimate of 15% based on data from the Ischemic Optic NeuropathyDecompression Trial.²⁸ In contrast, recurrentAION in a previously affected eye is extremely uncommon. Only 1 of 21 eyeswith greater than a 6-month follow-up experienced a second episode of AION,occurring 10 years after the first event. Recurrent NAION has been reportedin fewer than 5% of patients.²⁷

Based on these observations, many of the features of AION associatedwith ODD are strikingly similar to those of standard NAION. Specifically,the prevalence of vascular risk factors, the patterns of visual field loss,and the frequency of progressive visual loss are all comparable in the 2 groups.This suggests that the pathogenesis of visual loss in DAION may be similarto that of NAION. Is there a common denominator in these 2 conditions thatmight explain such a similarity? Both conditions have been associated witha small scleral canal.^30,31 Thisanatomical configuration has been accepted as a factor in the pathogenesisof NAION because of axonal crowding with secondary vascular compromise.^32-34 Such crowding maysimilarly predispose to AION in eyes with drusen.

In contrast to the above similarities, we found some differences inour patient group compared with previously reported series of patients withNAION. While the age range of our patients was broad, the mean of 49.4 yearswas considerably lower than the 63.4 years reported for patients with NAION.One possible explanation for this difference is that the drusen bodies themselvesact directly on healthy optic disc vessels to cause infarction. Another possibleexplanation for this age difference might be related to the progressive thinningof the nerve fiber layer that occurs in eyes with ODD. Over time, the opticdisc becomes less crowded and, thus, may no longer be susceptible to discinfarction.³⁵ In other words, if a patientwith ODD reaches the age of 60 years without experiencing a disc infarction,the likelihood of doing so thereafter may be small. In support of this explanation,we found that only 3 of our 20 patients with ODD experienced an episode ofAION after the age of 60 years. Finally, the age difference could reflecta sampling bias because of more frequent imaging studies in young patientswith AION.

Another contrast between our patient group and patients with NAION withoutdrusen is the difference in overall visual outcome. Almost two thirds of involvedeyes in our series had a final visual acuity of 20/40 or better, and only3 were left with a visual acuity worse than 20/200. This more favorable outcomemay reflect the younger age of our patient group. The optic disc in a youngerindividual may have more capacity for autoregulation and collateral circulationthan that of an older patient. In support of this concept, we found that our3 youngest patients, aged 18, 26, and 38 years, had final visual acuitiesof 20/30, 20/30, and 20/400 in the involved eye, respectively, compared withour 3 oldest patients, aged 62, 65, and 69 years, who had final visual acuitiesof 20/60, 20/400, and 20/50 in their involved eye, respectively. The visualacuities at onset were also better in our series compared with those of NAIONpatients without drusen, but such differences among studies are difficultto interpret because of variability in the timing of the initial examination.

We also found a greater frequency of transient obscurations of visionin our patients compared with those with standard NAION. While transient visualobscurations are well described in patients with ODD,³⁶ theyare quite uncommon in patients with NAION. A history of transient visual obscurationsin a patient with NAION should, thus, raise the possibility of underlyingdrusen. Although we noted a higher frequency of pain in our group, the discomfortin all patients was mild, described as a pressure or an ache rather than actualpain, and was self-limited in all. The significance of this observation isunclear. We considered the possible alternative diagnosis of optic neuritisin these individuals, but this seemed unlikely on clinical grounds.

In summary, our study supports a real (noncoincidental) associationbetween ODD and AION. Patients with DAION are generally younger than 50 years,younger on average than patients with standard NAION. In most other respects,their clinical features are quite similar to those of patients with standardNAION, including the prevalence of vascular risk factors, the pattern of visualfield loss, and the occurrence of a similar event in the fellow eye. In mostpatients, the visual outcome is more favorable than in patients with NAIONwithout drusen, and the likelihood of a recurrent event in an affected eye,as in patients with standard NAION, is low. Patients can be counseled accordingly.

This condition may be more common than previously appreciated. The opticdisc in the fellow eye of a patient with NAION typically appears crowded andfull. In some individuals, this appearance may reflect underlying (buried)disc drusen rather than simply the typical NAION disc at risk. Drusen thatare located more posteriorly (near the lamina cribrosa) may be more likelyto cause visual loss than those on the disc surface.³⁷ Itis these deep drusen that are more difficult to detect on an ophthalmoscopicexamination. Moreover, the disc elevation characteristic of buried drusenin young individuals is less prominent in older patients in whom progressiveaxonal dropout has occurred, thus making the diagnosis of drusen more challenging.A prospective study using ultrasonography in patients with NAION would furnisha better assessment of the actual frequency of drusen in patients with thisdisorder.

Corresponding author and reprints: Valerie Purvin, MD, Midwest EyeInstitute, 201 Pennsylvania Pkwy, Indianapolis, IN 46280 (e-mail: vpurvin@iupui.edu).

Submitted for publication March 27, 2003; final revision received July30, 2003; accepted August 25, 2003.

This study was supported in part by the Midwest Eye Foundation, Inc,Indianapolis, Ind (Department of Ophthalmology, Indiana University Schoolof Medicine, Indianapolis); and by an unrestricted development grant fromResearch to Prevent Blindness, Inc, New York, NY (Department of Ophthalmology,Indiana University School of Medicine).