Journal of Neuro-Ophthalmology, March 1997, Volume 17, Issue 1

Neuro Ophthalmology of the Pregeniculate Afferent Visual System

Journal of Neiira- Ophthalnmhxy 17( 1): 60- 77, 1997. © 1997 LippincoII- Raven Publishers, Philadelphia Annual Review Neuro- Ophthalmology of the Pregeniculate Afferent Visual System Developments in 1996 ( Part I) Laura J. Balcer, M. D., David R. Lynch, M. D., Ph. D., and Steven L. Galetta, M. D. EVALUATION OF THE AFFERENT VISUAL SYSTEM This year, 1996, was marked by continued advances in neuroimaging and the application of existing technolo­gies to new disorders. Neuroimaging techniques continue to be essential to the accurate and complete evaluation of disease processes involving the afferent visual system ( 1- 3). In a recent review, Davis and Newman ( 3) em­phasized the importance and relevance of such tech­niques as magnetic resonance imaging ( MR1), magnetic resonance angiography ( MRA), computed tomography ( CT), ocular sonography, and conventional contrast an­giography. From the perspectives of a ncuro-ophthalmologist and a radiologist, they discussed appro­priate selection of imaging modalities based on the pathophysiologic nature and localization of the suspected disease process. Described by the authors as a " practical review for the ophthalmologist," this article provides an ideal reference for all who evaluate patients with visual pathway disease. A review by Villablanca et al. ( 4) like­wise provided a valuable overview of the neuroimaging studies used in the diagnosis of disorders that affect the afferent visual system. As emphasized by Davis and Newman ( 3), there has been a general trend toward increased use of noninvasive neuroimaging modalities, such as MRA, although con­ventional angiography remains the definitive technique for the detection of intracranial aneurysms. This concept was reinforced by Weinberg ( 5), who reported a 76- year- Manuscripl received 12/ 10/ 96; accepted 1/ 30/ 97 From ( he Division of Neuro- Ophthalmology, Departments of Neu­rology and Ophthalmology, Hospital of the University of Pennsylvania, Scheie Eye Institute, University of Pennsylvania School of Medicine Address correspondence and reprint requests to Dr. Steven L. Gal­etta, Department of Neurology, 3 West Gates, 3400 Spruce St., Phila­delphia, PA 19104, U. S. A. old patient with a painful, incomplete, pupil- involving third nerve palsy. An aneurysm was strongly suspected. Although an MRI and MRA were initially unrevealing, conventional cerebral angiography demonstrated a 10- mm aneurysm at the junction of the posterior communi­cating and internal carotid arteries. The author and com­mentators emphasized that, retrospectively, the aneu­rysm was present on the original MRA and that it was not well visualized due to the fact that small subvolume (" slice- by- slice") analysis was not performed. Their case illustrates that MRA has the potential to be useful in the future for the detection of aneurysms. However, until acquisition and interpretation techniques for MRA are further refined and established by large well- performed studies, conventional angiography remains the imaging modality of choice. Moster et al. ( 6) demonstrated that functional brain imaging techniques, including positron emission tomog­raphy ( PET) and single proton emission tomography ( SPECT), may be complementary to MRI and CT for localizing visual loss in the retrochiasmal pathways. They reported two patients with a history of anoxic brain injury, binocular visual loss, and normal pupillary reac­tions. CT and MRI were normal in one patient, and in the second patient MRI showed subtle signal abnormality in the occipital gray matter. Both patients were initially thought to have functional visual loss on the basis of unremarkable imaging studies; however, SPECT scan­ning in the first case demonstrated decreased perfusion in the posterior occipital lobes, and PET scanning in the second case showed markedly decreased glucose me­tabolism in the visual association areas. These cases em­phasize the value of functional imaging techniques for investigating binocular visual loss in the setting of an­oxic brain injury, degenerative disease, and other condi­tions for which structural imaging studies may be unre­vealing. Miki et al. ( 7) examined the relationship between vi- 60 PREGENICULATE AFFERENT VISUAL SYSTEM 51 sual field defects related to optic nerve or chiasmal dis­orders and patterns of visual cortex activation by func­tional MRI. Seven patients with anterior visual pathway field defects secondary to traumatic or compressive optic neuropathy, glaucoma, or chiasmal compression under­went ophthalmologic examination, automatic perimetry, and functional MRI. Three normal control subjects were similar evaluated. The two patients with chiasmal com­pression had bitemporal hemianopsia; the remaining five patients had unilateral visual field defects and optic at­rophy. Functional MRI in the normal controls demon­strated similar levels of primary visual cortex activation with monocular checkerboard stimulation of each eye. The patients with unilateral optic neuropathy had rela­tively reduced visual cortex activation when the affected eye was stimulated, whereas those with bitemporal hemi­anopsia had reduced or absent activation of visual cortex contralateral to the visual field defect for each eye. This study demonstrated that functional MRI may not only be useful for the assessment of primary visual cortex pa­thology, but may be of potential experimental and clini­cal value for the investigation of anterior visual pathway disorders ( 7). For detecting and localizing visual field defects result­ing from lesions of the anterior and posterior visual path­ways, automated perimetry is the most widely used tech­nique. This method is also used for serial follow- up of patients with ophthalmic and neuro- ophthalmic disease. An ophthalmic procedure assessment published recently by the American Academy of Ophthalmology ( 8) dis­cussed the types and standard uses of automated perim­etry in clinical practice. The role of suprathreshold static perimetry as a rapid and effective method for detecting visual loss in patients with neuro- ophthalmologic disease was evaluated by Siatkowski et al. ( 9). They performed central 30° automated static full- threshold perimetry and suprathreshold perimetry on 159 consecutive patients, half of whom had true neuro- ophthalmologic disease. Sensitivities and specificities for detecting true visual field defects were comparable between the two tech­niques. However, the mean amount of time required for the full- threshold examination was 14.8 min, compared with 3.5 min for suprathreshold perimetry. This differ­ence in time may be critical in obtaining an accurate field in the neurologically impaired patient. The authors em­phasize that full- threshold testing is still necessary for detailed quantification of visual field defects once their presence is established and for subsequent follow- up ex­aminations. This may be particularly important in disor­ders associated with progressive optic nerve dysfunction such as glaucoma and papilledema. Automated and manual perimetry are frequently used to examine patients with functional visual loss. The char­acteristics and patterns of field abnormalities, particu­larly those demonstrated on a Goldmann perimeter, are often used to support clinical suspicions of functional versus organic visual loss. In a very interesting study of functional and fake perimetry, Thompson et al. ( 10) demonstrated that, with minimal coaching, normal sub­jects were able to imitate physiologic visual field abnor­malities ( such as enlarged blind spots and quandrantic, hemianopic, and attitudinal defects) without difficulty on both manual and automated perimetry. Paracentral and cecocentral scotomas were more difficult to imitate and reproduce on automated perimetry. Perhaps the most in­teresting findings is the fact that, when performing manual perimetry, experienced technicians were less able to differentiate patients with functional visual loss from those with organic disease than their less experi­enced peers. The results of visual field testing are also subject lo the influences of previous patient experience with the lest, or perimetric learning. This was demonstrated in an inves­tigation by Heijl and Bcngtsson ( 11) of the effects of perimetric experience in patients with glaucoma. In that study, each of 25 patients with newly diagnosed glau­coma underwent five consecutive weekly examinations using full- threshold automated perimetry. Between the first and second test sessions, average mean deviation values improved by 2.81 dB ( p < 0.001), but did not change significantly thereafter. Learning curves demon­strating changes in mean deviation were steepest for the group of eyes with moderate visual field loss. Subjective improvement was " obvious" in 21 of the 37 eyes tested based on the gray- scale printouts. The results of this study have implications for clinical trials utilizing peri­metric techniques as well as for the serial follow- up of patients with neuro- ophthalmic disease. As indicated by the authors ( 11), baseline visual field testing for clinical and investigational purposes should include data from two or more sessions. In the setting of therapeutic trials, control subjects should be used to minimize the con­founding effects of perimetric learning. The potential for young children to undergo conven­tional automated perimetry examinations was investi­gated by Safran et al. ( 12). They studied 42 girls aged 5- 8 years, each of whom underwent three successive automated perimetry sessions following a familiarization program. With increasing age, the children were better able to perform instructions correctly, were more likely to perceive dim stimuli, and required less time to com­plete the examination. The authors comment that the rates of false positive and false negative answers were surprisingly low for all age groups. All children, with the exception of one 5- year- old girl, were able to complete the testing successfully. The results of this study suggest that in children as young as 5 years of age with relatively preserved visual function automated perimetry may rep­resent a feasible examination technique if adequate pre­test explanations are provided. An important technique for the measurement of visual function in even younger children is the Teller Acuity Card procedure. Using this procedure, estimates of grat­ing visual acuity can be obtained rapidly for infants and young children in the office setting. Such estimates, however, are dependent upon subjective judgments by the observer regarding the child's head and eye move­ments in response to gratings of various sizes. Get/, et al. ./ Neitro- Opluhalnml, Vol. 17. No. 1. 1997 62 L. J. BALCER ET AL. ( 13) examined the interobserver test- retest reliability of the Teller Acuity Card procedure in normal children and in those with ophthalmic or neurologic abnormalities. Fifty- seven children aged 3- 38 months were tested in each group by two independent observers. Similarly high rates of interobserver agreement were found in both the normal and abnormal patient groups, suggesting that the procedure provides a reliable measure of acuity for chil­dren with ophthalmic and neurologic disease. Measures of visual function used in the Optic Neuritis Treatment Trial were examined by Trobe et al. ( 14) for their correlation with and sensitivity as indicators of vi­sual dysfunction. The authors found that additional test­ing beyond Snellen visual acuity, including Pelli- Robson contrast sensitivity, Humphrey visual field mean devia­tion, and Farnsworth- Munsell 100- hue color vision test­ing, would be of clinical value in the detection of optic neuropathy. A normal result for any one of these three nonacuity tests was found to be predictive of normal visual acuity at 6 months after onset. The Pelli- Robson contrast sensitivity test was found to be the most sensi­tive indicator of visual dysfunction in the setting of nor­mal visual acuity; this observation was most significant at 6 months after the baseline measurement. The results of this study are consistent with previous observations that contrast sensitivity is a sensitive measure of visual dysfunction in optic neuritis and is likely to be useful in the evaluation of other optic neuropathies. Two other clinical measures of visual function that have received recent comments in the literature include the duochrome test ( 15) and the Worth 4- dot neutraliza­tion test ( 16). Ing et al. ( 15) outlined the proper use of the duochrome colored lens test for the detection of func­tional visual loss in the office setting. The relationship between measurements of afferent pupillary defect ( APD) and recovery of fusion on the Worth 4- dot and Titmus stereo tests using neutral density filters was ex­amined in a recent study by Johnson ( 16). He observed a significant positive correlation between measurements of APD and the log filter density required for recovery of fusion on the Worth 4- dot test; however, no improve­ment of stereopsis was found. Arnold et al. ( 17) demonstrated that fluorescein angi­ography may be a useful technique for the differentiation of ischemic and nonischemic optic disc edema. They compared fluorescein angiograms from 22 patients with nonischemic optic disc swelling, including acute papil­litis, papilledemia, and intraorbital compression, with those of age- matched controls and of patients with non-arteritic anterior ischemic neuropathy ( NAION). A sig­nificant delay of prelaminar optic disc filling of at least 5 s was observed in 76% of patients with NAION ( 62% of those with NAION aged 46 and younger), but was not found in any of the patients with nonischemic disc swell­ing. Such a delay in prelaminar optic disc filling was thus found to be highly specific for ischemic disc swelling in this group of patients. For the evaluation of blood flow within the eye and orbit, color doppler ultrasound imaging is a useful non­invasive technique ( 18). Williamson and Harris ( 18) dis­cussed the utility of this technique in the diagnosis of central retinal artery and vein occlusions, nonarteritic ischemic optic neuropathy, intraocular tumor, glaucoma, diabetic retinopathy, carotid cavernous fistula, and or­bital apex tumors. To enable evaluation and serial follow- up of retinal pathology, a computer- based algorithm has been devel­oped by Mahurkar et al. ( 19) for constructing 100° field-of- view composite fundus images. The authors ( 19) sug­gest that this method should reduce the factors of mis­alignment and variations in color and illumination, making the assessment of changes in the peripheral retina less difficult. NEURO- OPHTHALMOLOGY AND THE RETINA From a basic science perspective, the retina has re­cently been the focus of numerous interesting investiga­tions, the results of which are relevant to the study of normal brain function and neurodegeneration. That a 24- h circadian clock is present in mammalian retina was reported in Science ( 20,21). Melatonin synthesis was found to occur in a circadian fashion in cultured hamster retinas exposed to light cycles in vitro. This circadian rhythmicity of melatonin synthesis is likely essential for normal photoreceptor function, including the regulation of replacement of rod photoreceptors ( 20,21). Conse­quently, a further understanding of this mechanism may have implications for the investigation of retinal degen­erative diseases ( 20,21). Neurotransmitters have important roles in the patho­genesis of retinal and other CNS degenerative diseases. Elevated levels of the excitatory neurotransmitter glutamate have been demonstrated within vitreous sam­ples from humans and monkeys with glaucoma ( 22). Drey-er et al. ( 22) found that glutamate levels in the vitreous of patients with known glaucoma and inadvertent posterior capsule rupture during cataract extraction were twice as high as those of patients with cataracts only. A similar result, with an even move pronounced relative elevation of glutamate levels, was obtained when vitreous samples from monkeys with and without glaucoma were com­pared. The results of this investigation indicate a poten­tial excitotoxic role for glutamate in the damage of reti­nal ganglion and optic nerve cells in glaucoma ( 22). That chronic exposure to glutamate in low concentra­tions is toxic to retinal ganglion cells was further dem­onstrated by Vorwerk et al. ( 23). Excitotoxicity through glutamate receptors has been demonstrated in essentially all areas of the CNS. Such toxicity may result from brief exposures to high levels of glutamate or to longer expo­sures of lower levels of glutamate or its analogues. This study ( 23) demonstrated that retinal ganglion cells are sensitive to low levels of glutamate, thus suggesting glutamatergic exposure as a mechanism in chronic optic neuropathies such as glaucoma. In addition, this toxicity is mediated by a specific glutamate receptor, the N-methyl- D- aspartate receptor. Pharmacologic antagonists of this receptor include phencyclidine ( PCP), dex- ./ Neuro- Opluhalmol, Vol. 17. No. I. 1997 PREGENICULATE AFFERENT VISUAL SYSTEM 63 tromethorphan, and the antiparkinsonian agent meman-tine, and a host of compounds currently under develop­ment. These agents may prove useful in chronic therapy of optic neuropathies such as glaucoma. Studies of the effects of aging on the structural and functional aspects of the neuroretina may also provide insight into retinal and CNS degenerative disease. The effects of aging on the densities, numbers, and sizes of retinal ganglion cells in the Rhesus monkey was inves­tigated by Kim et al. ( 24). Flat mounts of retina from each of seven young adult and six old rhesus monkeys were analyzed with respect to the total number of gan­glion cells within each retina, the densities of ganglion cells in each of four retinal quadrants, and the soma sizes of ganglion cells. No significant differences were found between the young and old retinas for any of these pa­rameters. According to the authors ( 24), the results of this study corroborate those of previous studies of the lateral geniculate nucleus ( LGN) and suggest that, in rhesus monkey, the anterior visual pathways are not sig­nificantly affected structurally by the process of aging alone. However, aging may have an effect on the density of retinal pigment epithelial ( RPE) cells, according to an­other recent study ( 25). In that investigation of 53 normal human donor eyes from patients aged 18- 85 years, the density of retinal pigment epithelial cells in the fovea was found to decrease by ~ 0.3%/ year with increasing age. Although loss of RPE cells with age was most ap­parent in the fovea, other regions of the retina were com­parably affected, and the correlation with age was like­wise statistically significant. The authors emphasize that the loss of RPE cells with age parallels that of photore­ceptors and optic nerve fibers, indicating that the aging process itself may affect the timing and pace of progres­sion of degenerative diseases, including open angle glau­coma ( 25). In addition to the RPE cell and photoreceptor cell loss associated with aging, an investigation by Bar-reau et al. ( 26) demonstrated that deletions in mitochon­drial DNA ( mtDNA) within human RPE and neuroretina cells also accumulate with age. That pathologic changes in the neuroretina may reflect the presence of degenerative disease elsewhere in the CNS is illustrated by the results of a two- part investiga­tion on retinal pathology in Alzheimer's disease ( 27,28). In light of previous reports indicating that patients with Alzheimer's disease exhibit signs of visual impairment and that retinal ganglion cell degeneration and optic neu­ropathy may occur, Blanks et al. ( 27,28) performed ex­periments to determine the extent of foveal/ parafoveal ganglion cell loss, regional neuronal loss, and glial cell changes in the retinas of patients with autopsy- confirmed Alzheimer's disease and those of age- matched control patients. Numbers of ganglion cells in the foveal and parafoveal retina were significantly reduced, by 25%, in patients with Alzheimer's disease ( 27). Examination of the ganglion cell layer of the entire retina likewise re­vealed significant loss of neurons ( by 36.4%, p < 0.004) and a significantly higher astrocyte to neuron ratio ( by 82%, p < 0.0008) in Alzheimer's disease patients com­pared with controls ( 28). These results confirm patho­logic involvement of the retina in Alzheimer's disease. Another study by Hedges et al. ( 29) adds to the present evidence for retinal ganglion cell degeneration in pa­tients with Alzheimer's disease. Retinal photographs were analyzed for nerve fiber layer abnormalities by two independent, masked observers. Although retinal nerve fiber layer abnormalities were observed at a higher fre­quency in Alzheimer's disease patients, variability in the quality and interpretation of the photographs between observers may, according to the authors ( 29), limit the current clinical usefulness of this technique in the evalu­ation of patients with Alzheimer's disease. Nonetheless, this study adds to the increasing evidence that Alzhei­mer's disease affects the anterior as well as posterior visual pathways. The effects of cancer and cancer therapy on retinal function were discussed in several reports ( 30- 35). With respect to primary neoplasms involving the retina, Lee et al. ( 30) described a rare case of solitary retinal astrocy­toma in a 3- year- old child without optic nerve involve­ment. The authors emphasized that, in general, astrocytic neoplasms involving ocular structures are particularly rare in the absence of neurofibromatosis or tuberous scle­rosis ( 30). Retinal dysfunction as a manifestation of distant neo­plastic disease has also been the subject of several re­ports ( 31- 36). Melanoma- associated retinopathy ( MAR) is characterized clinically by night blindness, abnormal electroretinographic features consistent with loss of bi­polar rod function, and selective binding of antibody to rod bipolar cells ( 31). In a study of Wolf and Arden ( 3 1), retinal dysfunction in the setting of MAR was found by psychophysical computer- based methods to be selective for neurons within the magnocellular system. Such se­lective loss of neuronal function was not observed in patients with congenital stationary nightblindness, a con­dition characterized by rod dysfunction and similarly ab­normal electroretinography ( ERG) findings. That diagnoses other than paraneoplastic retinopathy must be considered in patients with cancer- associated visual loss was emphasized by Ing et al. ( 32). They re­ported a 58- year- old man with large cell lung carcinoma and rapidly progressive bilateral visual loss. The patient had a normal fundus examination, ERG, and MRI of the brain and orbits; however, a lumbar puncture revealed carcinomatous meningitis. The differential diagnosis of visual loss in the setting of cancer ( including carcinoma­tous meningitis), cancer- associated retinopathy ( CAR), was reviewed, and diagnostic and management issues were discussed. Ing et al. ( 32) also indicated that chemotherapy-induced retinopathy is in the differential diagnosis of cancer- associated visual loss. Cunningham et al. ( 33) re­ported a case of retinal microvasculopathy in a 26- year- old woman with bilateral scotomata 2 years follow­ing autologous bone marrow transplantation for acute promyelocytic leukemia. The patient had not received ./ Neitro- Oplilhalmol, Vol. 17. No. I. 1997 64 L. J. BALCER ET AE. radiation therapy; the authors suggested that chemo-therapeutic agents, even in the absence of radiation or cyclosporinc therapy, may cause toxic retinal micro-vasculopathy. The autoimmune basis for cancer- associated retinopa­thies was discussed by Thurkill ( 34) in a review of lung cancer- induced blindness and by Suzuki et al. ( 35) in a report of a patient with CAR and presumed retinal vas­culitis. In the latter report, fluorescein angiography was performed in a patient with small- cell lung carcinoma and positive antiretinal antibody testing ( 35). This dem­onstrated diffuse staining of retinal vessels, which, ac­cording to the authors ( 35), may suggest the presence of an autoimmune retinal vasculitis. They emphasized the need for further pathologic study of retinal vasculature in patients with CAR. Another systemic disorder, mastocytosis, has recently been associated with retinal dysfunction in one patient ( 36). Lesser et al. ( 36) described a patient with nyctalo­pia and systemic mastocytosis involving the skin, liver, and gastrointestinal tract. Visual field testing revealed bilateral cecocentral scotomas, and ERG demonstrated rod- cone deficiency. Vitamin A deficiency secondary to malabsorption was demonstrated to be the probable cause of the nyctalopia; treatments with vitamin A re­sulted in resolution of the visual field defects and ERG abnormalities. The authors ( 36) added mastocytosis to the list of systemic disease entities that may lead to vi­tamin A deficiency and resultant nyctalopia. Primary retinal vasculitis in the absence of an associ­ated systemic etiology was the subject of a review by George et al. ( 37). They examined retrospectively the records of 25 patients with primary retinal vasculitis to determine whether the vasculitis was a manifestation of underlying systemic disease. None of the patients had evidence of a systemic inflammatory disease by history or serologic testing at the time of presentation. In the absence of signs or symptoms suggestive of an underly­ing systemic disease, the authors recommended that a minimal laboratory evaluation, including fluorescein an­giography, complete blood count, erythrocyte sedimen­tation rate, urinalysis, fluorescent treporemal antibody absorption test ( FTA- ABS), rapid plasma region ( RPR), chest roentgenogram, and, as appropriate, human immu­nodeficiency virus ( HIV) serology. Additional testing should be guided by the results of these studies and by suggestive findings in the history or examination. Findings of retinal venous sheathing and disruption of the blood- retinal barrier in patients with multiple sclero­sis ( MS) was examined in a 6- month longitudinal study by Birch et al. ( 38). Bilateral retinal venous sheathing was noted in six of the 23 patients. Of those six patients, three had venous leakage by fundus fluorescein angiog­raphy ( FFA) associated with the sites of venous sheath­ing. The venous sheathing remained unchanged through­out the 6- month follow- up period, and the venous leak­age likewise persisted in two patients. In one patient, venous leakage appeared after 3 months of follow- up and resolved during the next 3 months. However, in this group of patients, the presence of retinal venous inflam­matory changes ( sheathing) and apparent disruption of the blood- retinal barrier did not correlate with the num­bers of new gadolinium- enhancing brain lesions detected by MRI. Clinical classifications of MS ( i. e., relapsing-remitting versus chronic progressive) and numbers of clinical relapses during the 6- month period also did not correlate with the presence or absence of retinal findings. The authors ( 38) suggested that these retinal findings, though consistent with blood- retinal barrier disruption, occur infrequently in MS and, like the brain MRI find­ings, do not correlate with clinical course of the disease. Recurrent CNS inflammation in association with acute zonal occult outer retinopathy ( AZOOR) was reported in one patient by Jacobson ( 39). This 28- year- old woman had characteristic clinical and electrophysiologic fea­tures of AZOOR, but also had CSF lymphocytosis and multiple areas of increased signal intensity on brain MRI. No other systemic abnormalities were found. The retinal symptoms and abnormalities were recurrent. Six years following her initial presentation, the patient developed an acute cervical myelopathy with CSF lymphocytosis. Jacobson ( 39) emphasized that, although the clinical spectrum of AZOOR may include CNS inflammation, this patient may have had CNS vasculitis or demyelina-tion independent of the retinal pathology. Infectious disorders of the retina associated with HIV infection and acquired immunodeficiency syndrome ( AIDS) have been the subject of many recent investiga­tions and reviews ( 40- 64). Several editorials ( 4CM- 3) ad­dressed the spectrum of HIV- related eye disease and fo­cused on the importance of early recognition and treat­ment to the preservation of vision. Kupfer ( 42) emphasized the role of randomized clinical trials in the achievement of these goals. MacLean et al. ( 43) in Aus­tralia reviewed the records of 723 patients with AIDS who had ophthalmic consultations. They found that the most common abnormalities were microvasculopathy with cotton wool spots and hemorrhages associated with early HIV infection, external disease ( including Kaposi's sarcoma), and cytomegalovirus ( CMV) retinitis ( present in 204 of the 723 patients). Many reports this year have examined the epidemiol­ogy, pathophysiology, clinical assessment, and treatment of CMV retinitis in patients with HIV infection ( 44- 61). Hoover et al. ( 44) investigated the incidence and preva­lence of CMV retinitis in a cohort of HIV- infected ho­mosexual men with CD4 counts of < 100/| xl. Of the 367 patients followed for a 4- year period, 103 patients de­veloped CMV infection and 73 ( 71 %) of these had ret­initis. Thus, the probability of developing CMV retinitis during the 4- year period following the first CD4 count of < 100/ uJ was 25% for those patients. Of those who de­veloped CMV retinitis, 81% did so after the CD4 count had fallen to 50/ pT The authors state that these results support the accepted clinical practice of periodic ocular examinations for patients with CD4 counts of < 100/| JL1 and intensified screening once the count is < 50/| xl. In addition to the more global aspects of CMV retini- ,/ Newo- Ophllmlmol, Vol. 17, No. I. 1997 PREGENICULATE AFFERENT VISUAL SYSTEM 65 tis, an understanding of replication of the virus at the cellular level is also necessary for the further develop­ment of rational therapies. Detrick et al. ( 45) investigated the ability of human CMV to replicate in human retinal pigment epithelial ( RPE) cells in comparison with other cell types in vitro. This study demonstrated that CMV was able to replicate in cultured RPE cells, despite the fact that the virus typically does not replicate in epithelial cells in vitro. Furthermore, a distinct pattern of infection was observed in RPE cells, including a prolonged period of viral production prior to the onset of cytopathology, as well as a unique pattern of viral protein expression. The authors indicated that these findings may provide expla­nations for viral persistence within the retina and latency of infection ( 45). That disruption of the blood- retinal barrier may be a factor in the spread of CMV to the retina during systemic infection was investigated by Duan et al. ( 46). Using a BALB/ c mouse model, these authors found that nonim-munosuppressed mice had significantly higher viral titers in eyes in which the blood- retinal barrier had been dis­rupted, and titers for disrupted eyes were even higher in immunosuppressed mice. Most interesting, however, was the fact that, among the eyes of immunosuppressed mice, those in which the blood- retinal barrier had been dis­rupted demonstrated virus within both inner ( retina, RPE, iris, ciliary body) and outer ( conjunctiva, sclera) ocular structures. This result was in contrast to eyes of immunosuppressed mice without blood- retinal barrier disruption, in which virus was found only within outer ocular structures. Citing results from previous studies, the authors ( 46) suggested that disruption of the blood-retinal barrier may occur in humans by virtue of HIV infection of the retina and retinal vasculature, and that such disruption may provide one possible mechanism of retinal involvement. The diagnosis of CMV retinitis is generally based on clinical examination without confirmation from labora­tory testing. The value of polymerase chain reaction ( PCR) as a confirmatory technique in the diagnosis of CMV retinitis was examined by Doornenbal et al. ( 47). Undiluted aqueous or vitreous human samples from pa­tients with AIDS and clinically suspected CMV retinitis or uveitis were obtained. Using both PCR and antibody detection techniques, these samples were compared with those from a group of patients with uveitis of suspected nonherpetic etiology and a group of healthy controls. CMV DNA was detected by PCR in 10 of the 11 ocular fluid samples from the CMV retinitis group, whereas no CMV DNA was detected in the nonherpetic uveitis or healthy control groups. PCR was also found to be more sensitive than antibody detection techniques in this group of patients. The authors ( 47) suggested that further stud­ies are needed to evaluate the efficacy of PCR and to potentially develop protocols for its use. An accompa­nying editorial by Mitchell ( 48) stated that the conclu­sions of Doornenbal et al. ( 47) are consistent with those of recent investigations of PCR assays for CMV and that this technique may contribute to future therapeutic deci­sion making. New treatments for CMV retinitis are being active­ly developed and tested ( 49). This is necessitated by the fact that, since the emergence of HIV infection, an estimated 40,000 cases of CMV retinitis occur in the United States each year ( 49). A recent editorial by Free­man ( 49) provided a useful summary of current and fu­ture treatments for CMV retinitis, including opinions re­garding the advantages and disadvantages of various strategies. Findings of the Cytomegalovirus Retreatment Trial were reported in January 1996 ( 50). In this multicenter, randomized, controlled clinical trial, the investigators sought to determine whether patients with persistently active or recurrent CMV retinitis would be treated most effectively with a combination of intravenous foscarnet and ganciclovir or with either agent as monotherapy. The median time to progression of retinitis was significantly longer ( 4.3 months) in the combination group than in those receiving foscarnet ( 1.3 months) or ganciclovir ( 2.0 months) monotherapy ( p < 0.001). Combination therapy was therefore demonstrated to be the most ef­fective treatment in this group of patients and may offer a more effective alternative to monotherapy in the setting of persistent or relapsed CMV retinitis ( 50). That the virostatic mechanism of action of ganciclovir may permit human CMV DNA to be maintained within the retina for future replication was suggested by Burd et al. ( 51). Following treatment of cultured CMV- infected human retinal glial cells with gancyclovir, both pro­cessed CMV DNA and infectious virus were recovered. The authors suggested that this observation may corre­late with recurrence of CMV retinitis following cessation of therapy ( 51). In addition to the risk of recurrence, patients receiving ganciclovir or any intravenous therapy via indwelling catheter are at risk for bacterial sepsis and, as in a case reported by Tufail et al. ( 52), infectious ophthalmic complications. They ( 52) described a patient with CMV retinitis who developed Staphylococcus au­reus bacteremia and related endophthalmitis confirmed by vitreal cultures. Because of the risks involved in intravenous treat­ments, local therapies, including intravitreal injections ( 53- 55) and intravitreal implants ( 56,57), have been in­vestigated as both adjunctive and primary therapies. In a study of 11 vitreous samples from seven patients with CMV retinitis, Morlet et al. ( 53) demonstrated that the presence of virostatic levels of intravitreal ganciclovir for 7 days following high- dose injection may explain its effectiveness in the prevention of relapse. They admin­istered a 2- mg dose of ganciclovir intravitreally in seven patients, and intravitreal concentrations were measured at 24 and 72 h. In this small group of patients, the mean intravitreal concentration remained above the IDS0 for CMV for up to 7 days, suggesting that weekly mainte­nance treatments may provide effective control of CMV retinitis. Desatnik et al. ( 54) treated one patient successfully with intravitreal foscarnet and ganciclovir after he re­fused systemic therapy. The authors ( 54) indicated that this case raises the possibility that combined intravitreal ./ Neuro- Oplulwlmol. Vol. 17, No. 1, 1997 66 L. J. BALCER ET AL. injections may be an effective alternative for patients who do not tolerate or refuse systemic treatment. Intravitreal cidofivir ( HPMPC) was also recently evaluated by Rahhal et al. ( 55) as maintenance therapy for CMV retinitis. They treated 53 eyes of 35 patients with AIDS and CMV retinitis with 20 | xg of intravitreal cidofovir at 5- 6- week intervals. Of the 53 eyes studied, 29 had previously received systemic anti- CMV therapy. During a mean follow- up period of 15 weeks, all eyes with active retinitis demonstrated healing and only four eyes ( all in the previously treated group) had evidence of retinitis progression. Marx et al. ( 56) investigated the efficacy of the gan­ciclovir implant for treatment of recurrent CMV retinitis. The implant, consisting of a 6- mg pellet of ganciclovir, was inserted into 91 eyes of 70 patients with recurrent active CMV retinitis in the setting of AIDS. Of the 70 eyes that had at least 1 month of follow- up postopera­tively, 53 ( 76%) had a positive response by 1 month, as evidenced by the presence of inactive CMV retinitis. Furthermore, 64% of the eyes with a positive response at 1 month did not develop recurrent retinitis during the study period. Of those eyes with a recurrence, the median time to recurrence was 7 months. The authors pointed out that most of the patients enrolled in the study continued to receive systemic anti- CMV medications, so that the ganciclovir implant might be used as adjunctive therapy for recurrent CMV retinitis ( 56). In one ease ( 57), however, use of the ganciclovir im­plant exclusively was demonstrated to be effective for long- term maintenance treatment of CMV retinitis. The patient, reported by Duker et al. ( 57), was a 33- year- old man with AIDS and unilateral peripheral CMV retinitis. He first received intravenous ganciclovir for 2 weeks with positive clinical response; this was discontinued, and a ganciclovir implant was inserted. The retinitis ap­peared inactive by 4 weeks following treatment. Follow­ing a recurrence in the same eye 8 months later, a second implant was interested, with resolution once again of the retinitis. During the next 2 years, the implant was re­placed three more times but without evidence of further recurrence. The authors emphasized, however, that local therapy in the absence of systemic treatment may in­crease the risk of systemic CMV or occurrence of reti­nitis in the fellow eye ( 57). Although this patient did develop retinitis in the contralateral eye, which was treated with intravenous ganciclovir followed by im­plants, he did not have significant systemic symptoms at the time of the report. The use of long- term suppressive therapies for CMV has, however, been associated with the appearance of virus that is resistant to the available antiviral agents ( 58). Jabs et al. ( 58) investigated the prevalence of gan­ciclovir- and foscarnet- resistant CMV isolates in a sample of patients with AIDS and newly diagnosed CMV retinitis. Of the 49 patients enrolled in the study during a 13- month period, 78% had a positive blood or urine culture for CMV. The authors concluded that re­sistance represented an uncommon occurrence, as only 2% and 4% had samples containing isolates resistant to ganciclovir or foscarnet, respectively ( 58). A review of 76 cases from the Foscarnet- Ganciclovir CMV Retinitis Trial ( 59,60) indicated the usefulness of high- quality fundus photographs for the assessment of retinitis and prompt detection of progression. Fundus photographs were used in that study ( 59,60) to evaluate treatment response, and comparisons were made between photo assessments at the clinical site and a centralized grading center. The authors suggested that centralized grading of photographs may increase reliability of as­sessment and reduce risk of observer bias ( 60). Patients with CMV retinitis in the Foscarnet- Ganciclovir trial also participated in the development and testing of a questionnaire to assess visual function and quality- of- life issues related to treatment ( 61). The resulting 18- item self- administered scale was found to be both reliable and valid; as stated by the authors ( 61), additional evaluation of this measure will be necessary to determine its utility in further studies of CMV retinitis. Herpes simplex virus ( HSV) was implicated as a pos­sible sole cause of necrotizing retinitis in two patients with AIDS described by Cunningham et al. ( 62). Both patients presented with rapidly progressive vision loss and had fundus examinations consistent with necrotizing retinitis. A PCR- based assay and restriction analysis of vitreous samples were negative for CMV DNA in both patients. However, the same analysis did reveal evidence of HSV- 1 DNA in one patient and HSV- 2 DNA in the other. The specificity of this particular assay method was found to be very high, as 163 control vitreous samples were negative for HSV DNA. This article suggests HSV as a possible sole etiology for progressive retinitis in the setting of HIV infection. Although the sensitivity and specificity of combined PCR and restriction enzyme analysis has not been entirely established, these tech­niques may prove useful if the specificity is high ( 62). HIV- related retinal microangiopathy syndrome ( MAPS) was the topic of a letter by Thierfelder et al. ( 63). Their data from a sample of 125 patients with HIV infection suggested a possible role for the hepatitis C virus ( HCV) in MAPS. MAPS is characterized by cot­ton- wool spots and intraretinal hemorrhages, and repre­sents an immunocomplex vasculitis in the setting of hy­pergammaglobulinemia ( 63). Of the 125 patients exam­ined, 50 were HCV seropositive; of these 50 patients, 27 had MAPS. In contrast, of the 75 patients who were HCV seronegative, only five had evidence of MAPS. The significantly higher prevalence of MAPS in the HCV seropositive group suggests that HCV may in­crease the risk of MAPS in patients with HIV infection. A case reported by Walton et al. ( 64) illustrated that metastatic bacterial endophthalmitis should be consid­ered in the differential diagnosis of enlarging single or multiple whitish subretinal lesions in patients with AIDS, particularly in the setting of fever or other symptoms consistent with bacterial sepsis. The patient described was a 44- year- old man with AIDS who developed Staph­ylococcus aureus sepsis and died of severe pneumonia. ./ Nciiro- Oplulmlmnl, Vol. 17, No. 1. 1997 PREGEN1CULATE AFFERENT VISUAL SYSTEM 67 Autopsy examination of the left eye revealed a choroidal abscess with clusters of gram- positive cocci in the sub-retinal space, but no overlying vitreous inflammation. Grossly, the abscess appeared as a 2 x 3 mm raised white lesion in the superonasal retina. The authors ( 64) empha­sized the importance of a systemic work- up when such lesions are noted in the presence of suspicious systemic symptoms. In an HIV- negative patient, Fusarium endophthalmitis was described by Gabriele and Hutchins ( 65) as a com­plication of intravenous cocaine use. A 30- year- old man developed monocular vision loss over 1 week. Examina­tion of the affected eye was remarkable for keratic pre­cipitates, aqueous and vitreous cells, inflammatory sheathing of the veins and arterioles, and retinal infiltrates. A pars plana vitrectomy was performed, and vitreous cul­ture revealed Fusarium dimerium fungus. The patient's vision recovered completely following treatment with in-travitreal and intravenous amphotericin B. The author ( 65) recommend consideration of Fusarium species as potential ocular pathogens in intravenous drug abusers, even in the absence of an immunocompromised state. The retinal findings associated with subacute bacterial endocarditis were reviewed by Weissgold and Decker ( 66) in a report of a patient with ventricular false chorda. This otherwise healthy 40- ycar- old woman developed blurred vision 5 days following a molar tooth extraction. Ophthalmic examination revealed an intraretinal hemor­rhage in the fovea of the left eye and numerous white-centered retinal hemorrhages ( Roth's spots) in both eyes. A cardiac murmur was noted, and blood cultures were positive for alpha- hemolylic Streptococcus morbillorum. A transesophageal echocardiogram ( TEE) showed a re­gurgitant aortic valve and ventricular false chorda ( a con­genital structural anomaly) associated with vegetations. These findings were demonstrated extremely well by TEE; the importance of this technique to the evaluation of patients with eye findings suspicious for SBE was emphasized by the authors ( 66). Newsom et al. ( 67) reported a 10- year- old girl with cat- scratch ncuroretinitis in whom serologic testing was useful in establishing the diagnosis. One month prior to the onset of visual loss, the child has been scratched by her kitten; 2 weeks after that, she had developed a febrile illness. Examination of the affected eye revealed marked optic disc and peripapillary retinal edema, exudates, and a macular star. Antibody liters for Bartonella henselae, the causative agent, were elevated by enzyme immuno­assay testing. The authors ( 67) briefly reviewed the clini­cal characteristics and differential diagnosis of ncurore­tinitis, emphasizing the practicality, sensitivity, and specificity of serologic testing to distinguish cat- scratch disease from other entities. Vascular disorders of the retina diagnosed and treated by neuro- ophlhalmologists often have age as an impor­tant risk factor. To investigate the influence of age on retinal and optic nerve head perfusion, Groh et al. ( 68) examined 85 healthy volunteers, aged 21- 82 years, using scanning laser Doppler flowimetry and pulsed Doppler sonography. Using these techniques, they found that reti­nal capillary blood flow and central retinal artery flow velocity decreased significantly with increasing age of subjects ( by - 6- 1 1%/ decadc). However, measurements of blood flow in the optic nerve head microcirculation did not differ significantly with respect to age. These results are consistent with those of previous investiga­tions, but the authors added that further correlation with morphologic and blood flow studies will be necessary to determine the exact mechanisms of these age- related changes ( 68). Grunwald el al. ( 69) used bidirectional laser Doppler velocimetry to measure maximum erythrocyte velocity in the retinal circulations of patients with insulin-dependent diabetes mcllitus of < 4 years in duration. The 19 diabetic patients studied had no evidence of diabetic retinopathy; however, compared with that in age-matched controls, the average calculated total volumetric venous blood flow rate in the diabetic group was 12% ( p < 0.05) higher. Also observed was a statistically signifi­cant positive correlation between duration of diabetes and increased total volumetric blood flow rate. The au­thors ( 69) indicate that this mild increase in blood flow early in the course of disease may influence later devel­opment of diabetic retinal microangiopathy. A review of the literature by Brown and Jampol ( 70) implicated vascular endothelial cell hormonal products as possible modulators for vascular autoregulation in the retina and choroidal circulations. They discussed the pos­sible roles of vasodilators, including nitric oxide and prostaglandins, and vasoconstrictors, such as endolhelins and renin- angiotensin products, which may be produced by vascular endothelial cells in response to blood pres­sure, oxygen levels, and chemical stimuli. The authors concluded that further investigation of vascular endothe­lial physiology will be necessary for the development of potential pharmacologic therapies for retinal vascular disorders ( 70). Changes in retinal and carotid blood flow in response to serotonin and collagen- induced platelet activation were examined in a primate atherosclerosis model by Sobey et al. ( 71). They studied normal monkeys, mon­keys who had been fed an atherogenic diet for 34 months ( atherosclerotic monkeys), and atherosclerotic monkeys who were fed a normal diet ( regression monkeys) for 8- 9 months with respect to both functional and structural abnormalities of internal carotid and retinal arteries. Both serotonin and collagen produced an increase in carotid blood flow in normal monkeys and a significant decrease in carotid blood flow ( compared with normals) in the atherosclerotic group. In monkeys fed the regression diet, however, there was a small increase in carotid blood flow in response to both compounds that was signifi­cantly different from the response in atherosclerotic monkeys, suggesting a reversal toward the normal state. Mean decreases in mean internal carotid artery diameter in response to serotonin were noted in all three groups of monkeys but were greatest in the atherosclerotic mon­keys. Retinal blood flow was decreased by serotonin but ./ Nairo- Ophihalmol. Vol. 17, No. I, I'W 6< S' L. .1. BALCER ET AE. increased by collagen in normals; flow decreased signifi­cantly in the atherosclerotic group but demonstrated a return toward the normal response in regression mon­keys. Thus, the increases in vasoconstrictor response seen in alherosclerotics were at least partially reversed by 7- 14 months on the regression diet. Structural changes tended to lag behind the functional changes, however, as increases in the intimal area within the ca­rotid and ophthalmic arteries noted in atherosclerotic monkeys persisted in the regression group. As indicated in an accompanying editorial comment by Wei ( 17), some physiologic vascular abnormalities that occur sec­ondary to atherosclerosis may be reversible with changes in diet. The generalizability of these findings to humans will depend on the relative contributions of other physi­ologic or lifestyle factors in humans. Atherosclerosis of the internal carotid artery as a risk factor for central retinal artery occlusion was present in a case reported by Jumper and Morton ( 72). They de­scribed an 87- year- old man with severe right internal carotid artery stenosis detected by ultrasound who un­derwent a chiropractic neck manipulation. Within min­utes of the procedure, the patient noted sudden visual loss in the right eye. Ophthalmic examination revealed a macular cherry- red spot and multiple retinal emboli. MRA confirmed a critical stenosis of the right internal carotid. The authors ( 72) emphasize thai stroke in the setting of ehirpractic manipulation is uncommon, but recommend a cautious approach to patients with possible or known carotid disease. In addition to the carotid artery, the heart represents another potential source of emboli in patients with acute retinal arterial occlusion. A retrospective review by Sharma el al. ( 73) examined 100 patients with acute cen­tral or branch retinal artery occlusion who underwent cardiac evaluation and transthoracic echocardiography. Of the 47 patients with echocardiography abnormalities, 10 had findings that prompted anticoagulation or cardiac surgery. Nine of these 10 patients had cardiac risk factors by history; the one patient who was not classified as high risk was found to have an atrial myxoma. Echocardiog­raphy was abnormal in 64% of the 67 patients with no cardioembolic risk factors. However, 98% of those with one or more risk factors ( high risk) had abnormalities. Although patients in the high- risk category in this study were 25 times more likely to have undergone anticoagu­lation or surgery, the authors ( 73) cautioned that the use of transesophageal echocardiography may have detected additional abnormalities requiring such interventions in both groups of patients. They emphasized the importance of cardiac risk factors in increasing clinical suspicion of a cardiac source of emboli in patients with acute central or branch retinal artery occlusion. Giant cell arteritis is another important risk factor for retinal arterial occlusion and other causes of acute isch­emic monocular visual loss. Fineman et al. ( 74) reported a rare case of nonembolic branch retinal artery occlusion as the presenting feature of giant cell arteritis. The pa­tient was a 77- year- old woman who developed monoc­ular visual loss following a 2- week history of headache. Fundus examination revealed a superior branch retinal artery occlusion in the newly affected left eye and a previously documented hemiretinal vein occlusion in the right eye. She subsequently noted a further decline in vision in the left eye associated with the onset of scalp tenderness, jaw claudication, and weight loss. Sedimen­tation rate was 99 mm/ h, and a temporal artery biopsy was consistent with giant cell arteritis. Although rarely reported as initial manifestation of giant cell arteritis, branch retinal artery occlusion should prompt evaluation for this diagnosis in elderly patients ( 74). As confirmed by a recent case- control study ( 75), cen­tral retinal vein occlusion ( CRVO), like central retinal artery occlusion, has a number of risk factors. In this investigation, the Eye Disease Case- Control Study Group examined 258 patients with CRVO and 1,142 control patients during a 4'/ 2- year period. Using odds ratio calculations, significant associations with increased risk of CRVO were found for diabetes mellitus, hyper­tension, history of cardiovascular disease, antithrombin III level, glaucoma, and increased erythrocyte sedimen­tation rate ( in women only). Use of a multivariate logistic regression model confirmed a significant association with CRVO for hypertension, diabetes, glaucoma, and increased ESR in women, whereas use of estrogens in women, use of alcohol, and physical activity were asso­ciated with decreased risk. These results are consistent with the authors' ( 75) view that a close association be­tween certain cardiovascular risk factors and CRVO ex­ists. A population study in Australia by Mitchell et al. ( 76) likewise demonstrated significant associations with retinal vein occlusion for hypertension, stroke, angina, and glaucoma in patients over 48 years of age. Increased blood viscosity, activated protein C resis­tance, and other coagulation abnormalities were also in­vestigated as possible risk factors for CRVO in a case-control study ( 77). Williamson el al. ( 77) examined 69 patients with central and hemiretinal vein occlusion and 69 age- matched controls. Patients were further divided into subgroups for comparison of particular characteris­tics, such as activated protein C resistance. The age range for the patients and controls was 27- 84 years, with a mean of 64.2 years. Mean hematocrit- corrected and rela­tive blood viscosities were higher, but not significantly so, in the patients with CRVO than in the controls. How­ever, after 1 year of follow- up, the corrected and relative blood viscosities for the patients were significantly lower than they had been at the time of occlusion. Compared with controls, patients with CRVO had a significantly higher prevalence of activated protein C resistance as well as higher mean levels of antithrombin III, von Will-ebrand factor, and plasminogen activator inhibitor. The finding of increased levels of antithrombin III was noted to be inconsistent with previous results, but the authors suggested that this may serve as a compensatory mecha­nism. Observations regarding the other significant fac­tors, including activated protein C resistance, are consis­tent with a possible etiologic role for hypercoaguable state ( 77). J Nnim- Ophlhalnml, Vol. 17. No. 1. IW7 PREGENICVLATE AFFERENT VISUAL SYSTEM 69 A subsequent investigation by Larsson et al. ( 78) ad­dressed the issue of activated protein C resistance as a risk factor for CRVO in young adults. They analyzed blood samples from 31 patients aged 19- 48 years ( me­dian, 40 years) with a history of CRVO within the pre­vious 1- 6- year period. Of these patients, eight ( 26%) had evidence of activated protein C resistance. When only those patients < 45 years of age were considered, the percentage was even higher ( 36%). Compared with pre­viously reported prevalences of activated protein C re­sistance of 2- 7% in this age group, the percentage of patients with activated protein C resistance was signifi­cantly higher in this investigation ( 95% confidence in­terval for sample, 17- 59%). As indicated by the authors ( 78), this dominantly inherited disorder may be a very important and relatively common risk factor for CRVO in young patients. Essential thrombocytosis, a rare disorder of myelopro- Iiteration associated with systemic vascular occlusion, was reported in one patient with CRVO by Yoshizumi and Townsend- Pico ( 79). They described a 59- year- old man who developed sudden, painless monocular vision loss with retinal and fluorescein angiographic findings consistent with CRVO. The platelet count was 856,000/ mm3, and a bone marrow biopsy demonstrated increased cellularity with numerous atypical megakaryocytes. As­pirin and hydroxyurea therapies were given, with subse­quent marked reduction of the platelet count. He then developed two episodes of neovascular glaucoma; inter­estingly, the second episode followed a documented in­crease in the platelet count from 244,000- 400,000/ mm3 to 626,000/ mnr\ Thus, CRVO with subsequent neovas­cularization may be implicated as a possible ischemic complication of essential thrombocytopenia ( 79). Finally, Aritiirk et al. ( 80) performed a study of 17 patients aged 55- 73 years with CRVO and 41 patients aged 45- 72 years with branch retinal vein occlusion ( BRVO). The affected eyes were compared with unaf­fected eyes of the same patients and of age- matched controls with respect to ocular axial length, as deter­mined by A- scan ultrasonography. The controls had no refractive error. The mean axial length measurement for eyes with CRVO ( 22.25 mm) was significantly shorter than those of unaffected contralateral eyes ( p < 0.05) and control eyes ( p < 0.001). For the eyes with BRVO, the mean axial length was significantly shorter compared only with eyes of controls. Nonetheless, this study raises an interesting hypothesis concerning ocular length as an­other potential risk factor for retinal vein occlusion ( 80). Prognostic factors for retinal vein occlusion were evaluated by Glacet- Bernard et al. ( 81) in a study of 169 patients. Of the 175 affected eyes examined, CRVO was noted in 127 eyes ( including seven eyes with hemicentral vein occlusion), whereas 48 eyes demonstrated BRVO. Patients were aged 23- 87 years, with a mean of 58 years; 35 patients were < 50 years old. Moderate, marked, or massive retinal ischemia developed in 54% of initially nonischemic eyes with CRVO. Characteristics signifi­cantly associated with poor outcome ( as defined by vi­sual acuity of < 20/ 40 or 5= 100 optic disc- diameters as a measure of area on fluorescein angiography of ischemia) were older age; male sex; increased number of vascular, hematologic, and ocular risk factors; initial presence of retinal ischemia or cystoid macular edema; and baseline visual acuity of < 20/ 40. The importance of close follow-up was emphasized ( 81), especially for patients at higher risk for poor prognosis. Kupersmith et al. ( 82) examined the possible role of venous stasis as a primary mechanism for retinopathy and choroidal effusion in the setting of arteriovenous shunts ( AVS) involving the cavernous sinus. They ana­lyzed retrospectively the angiographic and clinical find­ings of 10 patients with retinopathy or choroidal effu­sion, and compared these with the findings of 10 patients with visual dysfunction secondary to AVS and elevated intraocular pressure. Although severe ipsilateral ophthal­mic vein or cavernous sinus thrombosis was noted in nine patients with retinopathy and seven patients with choroi­dal effusion, severe thrombosis was seen in only three of the patients with elevated intraocular pressure. As em­phasized in the report ( 82), the previously implicated mechanisms of arterialization of venous flow and arterial " steal" phenomenon may be less important causes of retinal or choroidal dysfunction than ophthalmic venous thrombosis in patients with cavernous sinus AVS. Goldberg et al. ( 83) demonstrated, consistent with pre­vious suggestions, that transvenous embolization of cav­ernous sinus- dural fistulas may provide a reasonable al­ternative to transarterial embolization when this tech­nique is unsuccessful or it not feasible. They reviewed the characteristics and outcomes for 10 consecutive pa­tients who had been treated using transvenous emboliza­tion via anterior orbitotomy. In nine patients, transarte­rial embolization had been attempted unsuccessfully; in one patient, it was the only technique used. The fact that visual acuity loss and symptoms of orbital congestion resolved in all 10 patients was emphasized as an encour­aging indication of the potential safety and benefit of this procedure in appropriate patients by well- trained clini­cians ( 83). THE OPTIC NERVE Vascular, inflammatory, infectious, toxic, nutritional, and hereditary disorders of the optic nerve were ad­dressed in two comprehensive review articles by New­man ( 84) and Miller ( 85). These emphasized important previous investigations that have influenced the diagno­sis and treatment of optic neuropathies. Some more re­cent studies and reports have added to this knowledge. Dolichoectasia of the internal carotid artery is a known but rare cause of compressive optic neuropathy. A report by Colapinto et al. ( 86) stressed consideration of this entity in patients with visual loss. They described a 34- year- old woman who developed painless monocular de­creased acuity and visual field constriction. Compression of the ipsilateral optic nerve by a dolichoectatic internal carotid artery was demonstrated by MRI and cerebral ./ Nfiiro- Ophlhalinol. Vol. 17, No. 1. 1997 70 L. J. BALCER ET AL. angiography. The visual loss resolved almost completely following pterional craniotomy and optic canal decom­pression ~ 4 weeks after onset. To investigate the possibility that mechanical com­pression by a normal intracranial internal carotid artery may lead to an otherwise unexplained optic neuropathy, Golnik et al. ( 87) performed a comparative study of anatomy as demonstrated on digitized MRI images. They examined consecutive patients with unexplained optic neuropathy, as defined by the presence of optic disc pal­lor without history of swelling, visual field defects, af­ferent pupillary defect, previously normal coronal MRI, and normal serologic testing. These cases were matched to 10 control patients by age. Coronal MRI images from cases and controls demonstrating the area 3 mm anterior lo the optic chiasm were digitized using an optical scan­ner. Optic nerve diameter, internal carotid artery flow void diameter, and distance between the optic nerve and internal carotid artery flow void was measured on the digitized images for the 10 affected eyes, for contralat­eral unaffected eyes, and for both eyes of the control patients. Despite the fact that affected optic nerves were of significantly smaller mean diameter than those of un­affected contralateral eyes and control eyes, the mean measured distance between the optic nerve and ipsilat-eral internal carotid artery flow void was also signifi­cantly less for the atrophic nerves. This result led to the authors' ( 87) hypothesis that neurovascular compression or contact similar to that implicated in trigeminal neu­ralgia may be a factor in unexplained optic neuropathy. REFERENCES 1. Lueck C. I. Investigation of visual loss: neuro- ophthalmology from a neurologist's perspective. ./ Neurol Neruosurg Psychiatry 1996; 60: 275- 80. 2. Galetta SL, Liu GT, Volpc NJ. Diagnostic tests in neuro-ophthalmology. Neurol CVm 1996; 14: 201- 22. 3. Davis PC, Newman NJ. Advances in neuroimaging of the visual pathways. Am J Ophthalmol 1996; 121: 690- 705. 4. Villablanca P, Cumin J, Arnold A, Lufkin R. Orbit and optic nerve. Top Magn Reson / making 1996; 8: 87- 1 10. 5. Weinberg DA. Negative MRI versus real disease. Surv Ophthalmol 1996; 40: 312- 9. 6. Moster ML, Galetta SL, Schatz NJ. Physiologic functional imaging in " functional" visual loss. Surv Ophthalmol 1996; 40: 395- 9. 7. Miki A, Nakajima T, Takagi M, Shirakashi M, Abe H. Detection of visual dysfunction in optic atrophy by functional magnetic reso­nance imaging during monocular stimulation. Am ./ Ophthalmol 1996; 122: 404- 15. 8. American Academy of Ophthalmology. Automated perimetry. Ophthalmology 1996; 103: 1144- 51. 9. Sialkowski RM, Lam BL, Anderson DR, Fcuer WJ, Halikman AB. Automated suprathreshold static perimetry screening for delecting neuro- ophthalmologic disease. Ophthalmology 1996; 103: 907- 17. 10. Thompson JC, Kosmorsky GS, Eillis BD. Fields of dreamers and dreamed- up fields: functional and fake perimetry. Ophthalmology 1996; 103: 117- 25. 11. I lei jl A, Bengtsson B. The effect of perimetric experience in pa­tients with glaucoma. Arch Ophthalmol 1996; 114: 19- 22. 12. Safran AB, Laffi GL, Bullinger A, et al. Feasibility of automated visual field examination in children between 5 and 8 years of age. Br J Ophthalmol 1996; 80: 515- 8. 13. Get/. LM, Dobson B, Luna B, Mash C. Interobserver reliability of the Teller acuity card procedure in pediatric patients. Invest Oph­thalmol Vis Sci 1996; 37: 180- 7. 14. Trobe . ID, Beck RW, Moke PS, Cleary PA. Contrast sensitivity and other vision tests in the Optic Neuritis Treatment Trial. Am J Ophthalmol 1996; 121: 547- 53. 15. Ing EB, Younge BR, Leavitt JA, Filzpatrick PJ. Functional visual loss and the duochrome test. Can J Ophthalmol 1996; 30: 35- 6. 16. Johnson LN. The relative afferent pupillary defect and a novel method of fusion recovery with the Worth 4- dol test. Arch Oph­thalmol 1996; 114: 171- 5. 17. Arnold AC, Badr MA, Hepler RS. Fluorescein angiography in nonischemic optic disc edema. Arch Ophthalmol 1996; I 14: 293- 8. 18. Williamson TH, Harris A. Color doppler ultrasound imaging of the eye and orbit. Surv Ophthalmol 1996; 40: 255- 67. 19. Mahurkar AA, Vivino MA, Trus BL, Kuehl EM, Datilcs MB, Kaiser- Kupfcr MI. Constructing retinal fundus photomontages: a new computer- based method. Invest Ophthalmol Vis Sci 1996; 37: 1675- 83. 20. Tosini G, Menaker M. Orcadian rhythms in cultured mammalian retina. Science 1996; 272: 419- 21. 21. Morcll V. A 24- hour circadian clock is found in the mammalian retina. Science 1996; 272: 349. 22. Drcyer EB, Zurakowski D, Schumer RA, Podos SM, Lipton SA. Elevated glutamate levels in the vitreous body of humans and monkeys with glaucoma. Arch Ophthalmol 1996; 114: 299- 305. 23. Vorwerk CK, Lipton SA, Zurakowski D, Hyman BT, Sabel BA, Dreyer EB. Chronic low- dose glutamate is toxic to retinal ganglion cells: toxicity blocked by memanline. Invest Ophthalmol Vis Sci 1996; 37: 1618- 24. 24. Kim CBY, Tom BW, Spear PD. Effects of aging on the densities, numbers, and sizes of retinal ganglion cells in rhesus monkey. Neurohiol Aging 1996; 17: 431 - 8 . 25. Panda- Jonas S, Jonas JB, Jakobczyk- Zmija M. Retinal pigment epithelial cell count, distribution, and correlations in normal hu­man eyes. Am J Ophthalmol 1996; 121: 181- 9. 26. Barreau E, Brossas JY, Courtois Y, Treton JA. Accumulation of mitochondrial DNA deletions in human retina during aging. Invest Ophthalmol Vis Sci 1996; 37: 384- 91. 27. Blanks J, Torigoe Y, Hinton DR, Blanks RHI. Retinal pathology in Alzheimer's disease 1: ganglion cell loss in foveal/ parafoveal retina. Neurohiol Aging 1996; 17: 377- 84. 28. Blanks J, SchmidLSY, Torigoe Y, Ponello KV, Hinton DR, Blanks RHI. Retinal pathology in Alzheimer's disease II: regional neuron loss and glial changes in GCL. Neurohiol Aging 1996; 17: 385- 95. 29. Hedges TR, Galves RP, Speigelman D, Barbas NR, Peli E, Yardley CJ. Retinal nerve fiber layer abnormalities in Alzheimer's disease. Acta Ophthalmol Scand 1996; 74: 271- 5. 30. Lee J, Harvey P, Finlay RD, Berry PJ. Solitary astrocytoma of the retina in a child. Br J Ophthalmol 1996; 80: 674- 5. 31. Wolf JE, Ardcn GB. Selective magnoeellular damage in mela­noma- associated retinopathy: comparison with congenital station­ary nightblindness. Vision Res 1996; 36: 2369- 79. / 32. Ing EB, Augsburgcr JJ, Eagle RC. Lung cancer vvifh visual loss. Surv Ophthalmol 1996; 40: 505- 10. / 33. Cunningham ET, Irvine AR, Rugo HS. Bone ( narrow transplanta­tion retinopathy in the absence of radiatior/ therapy. Am J Oph­thalmol 1996; 122: 268- 70. / 34. Thurkill CE. Lung cancer- induced blindness. Lung Cancer 1996; 14: 253- 64. 35. Suzuki T, Obara Y, Sato Y, Saito G, lehiwata T, Uchiyama T. Cancer- associated retinopathy with presumed vasculitis. Am J Ophthalmol 1996; 122: 125- 7. 36. Lesser RL, Brodie SE, Sugin SL. Mastocytosts- induccd nyctalo­pia. ./ Neuroophthalmol 1996; 16: 115- 9. 37. George RK, Walton C, Whitcup SM, Nussenblatt RB. Primary retinal vasculitis: systemic associations and diagnostic evaluation. Ophthalmology 1996; 103: 384- 9. 38. Birch MK, Barbosa S, Blumhardt LD, O'Brien C, Harding SP. Retinal venous sheathing and the blood- retinal barrier in multiple sclerosis. Arch Ophthalmol 1996; 114: 34- 9. 39. Jacobson DM. Acute zonal occult outer retinopathy and central nervous system inflammation../ Neuroophthalmol 1996; 16: 172- 7. 40. Jabs DA. Acquired immunodeficiency syndrome and the eye- 1996. Arch Ophthalmol 1996;! 14: 863- 6. ./ Ncuro- Ophlhalmol, Vol. 17, No. I, 1997 PREGENICULATE AFFERENT VISUAL SYSTEM 71 41. Courtrighl P. The challenge of HIV/ AIDS- rclated eye disease. Br J Ophthalmol 1996; 80: 496- 7. 42. Kupfer C. General principles for AIDS research. Arch Ophthalmol 1996; 114: 862. 43. Maclean H, Hall AJH, McCombe MF, Sandland AM. AIDS and the eye: a 10- year experience. Aust N Z J Ophthalmol 1996; 24: 61- 7. 44. Hoover DR, Peng Y, Saali A, el al. Occurrence of cytomegalovirus retinitis afler human immunodeficiency virus immunosuppression. Arch Ophthalmol 1996; 114: 821- 7. 45. Detrick B, Rhame J, Wang Y, Nagineni CN, Hooks JJ. Cytomeg­alovirus replication in human retinal pigment epithelial cells: al­tered expression of viral early proteins. Invest Ophthalmol Vis Sci 1996; 37: 814- 25. 46. Duan Y, Hernandez R, Pang L, Atherton SS. Spread of murine cytomegalovirus to inner ocular structures following disruption of the blood- retina barrier in immunosuppressed BALB/ c mice. In­vest Ophthalmol Vis Sci 1996; 37: 935^ t0. 47. Doornenbal P, Baarsma GS, Quint WGV, Kijlstra A, Rothbarlh PH, Nicsters HGM. Diagnostic assays in cytomegalovirus retinitis: detection of herpesvirus by simultaneous application of the poly­merase chain reaction and local antibody analysis on ocular fluid. Br.) Ophthalmol 1996; 80: 235- 40. 48. Mitchell S. Diagnostic assays in cytomegalovirus retinitis. Br J Ophthalmol 1996; 80: 195- 6. 49. Freeman WR. New developments in the treatment of CMV reti­nitis. Ophthalmology 1996; 103: 999- 1000. 50. The Studies of Ocular Complications of AIDS Research Group in Collaboration With the AIDS Clinical Trials Group. Combination foscarnet and ganciclovir therapy vs. monotherapy for the treat­ment of relapsed cytomegalovirus relinitis in patients with AIDS. Arch Ophthalmol 1996; 114: 23- 33. 51. Bind EM, Pulido JS, Puro DG, O'Brein WJ. Maintenance of imp­licative intermediates of ganciclovir- treated human cytomegalovi-rus- infectcd retinal cells. Arch Ophthalmol 1996; 1 14: 856- 61. 52. Tufail A, Weisz JM, Holland GN. Endogenous bacterial endoph­thalmitis as a complication of intravenous therapy for cytomega­lovirus retinopathy. Arch Ophthalmol 1996; 114: 879- 80. 53. Morlet N, Young S, Naidoo D, Graham G, Coronco MT. High-dose intravitreal ganciclovir injection provides a prolonged thera­peutic intraocular concentration. Br J Ophthalmol 1996; 80: 214- 6. 54. Desalnik HR, Foster RE, Lowder CY. Treatment of clinically re­sistant cytomegalovirus retinitis with combined intravitreal injec­tions of ganciclovir and foscarnet. Am .1 Ophthalmol 1996; 122: 121- 3. 55. Ralihal FM, Arevalo JF, Munguia D, el al. Intravitreal cidofovir for the maintenance treatment of cytomegalovirus retinitis. Ophthal­mology 1996; 103: 1078- 83. 56. Marx JL, Kapusla MA, Patel SS, et al. Use of the ganciclovir implant in the treatment of recurrent cytomegalovirus retinitis. Arch Ophthalmol 1996; I 14: 815- 20. 57. Duker JS, Ashton P, Davis JL, Keller R, Chuang E. Long- term, successful maintenance of bilateral cytomegalovirus retinitis using exclusively local therapy. Arch Ophthalmol 1996; 114: 881- 2. 58. Jabs DA, Dunn . IP, Enger C, et al. Cytomegalovirus retinitis and viral resistance: prevalence of resistance at diagnosis, 1994. Arch Ophthalmol 1996; 114: 809- 14. 59. The Studies of Ocular Complications of AIDS Research Group in Collaboration With the AIDS Clinical Trials Group. Clinical vs. photographic assessment of treatment of cytomegalovirus retinitis: foscarnct- ganciclovir cytomegalovirus relinitis trial report 8. Arch Ophthalmol 1996; 114: 848- 55. 60. The Studies of Ocular Complications of AIDS Research Group in Collaboration With the AIDS Clinical Trials Group. Assessment of cytomegalovirus retinitis: clinical evaluation vs. centralized grad­ing of fundus photographs. Arch Ophthalmol 1996; I 14: 791- 805. 61. Wit AW, Coleson LC, Holbrook J, Jabs DA, for the Studies of Ocular Complications of AIDS Research Group. Measuring visual function and quality of life in patients with cytomegalovirus reli­nitis. Arch Ophthalmol 1996; 114: 841- 7. 62. Cunningham ET, Short GA, Irvine AR, Duker JS, Margolis TP. Acquired immunodeficiency syndrome- associated herpes simplex virus retinitis. Arch Ophthalmol 1996; 114: 834^ 40. 63. Thierfclder S, Linnert D, Grehn F. Increased prevalence of HIV-related retinal microangiopathy syndrome in patients with hepatitis C. Arch Ophthalmol 1996; I 14: 899. 64. Walton RC, Wilson J, Chan CC. Metastastic choroidal abscess in the acquired immunodeficiency syndrome. Arch Ophthalmol 1996; 114: 880- 1. 65. Gabriele P, Hutchins RK. Fusarium endophthalmitis in an intra­venous drug abuser. Am J Ophthalmol 1996; 122: 1 19- 21. 66. Weissgold DJ, Decker P. I. Retinal hemorrhages from septic emboli in a patient with a ventricular false chorda. Am J Ophthalmol 1996; 122: 117- 9. 67. Newsom RW, Martin TJ, Wasilauskas B. Cat- scratch disease di­agnosed serologically using an enzyme immunoassay in a patient with ncuroretinitis. Arch Ophthalmol 1996; I 14: 493- 4. 68. Groh MJM, Michelson G, Langhans MJ, Harazny J. Influence of age on retinal and optic nerve head blood circulation. Ophthalmol­ogy 1996; 103: 529- 34. 69. Grunwalci JE, DuPont J, Riva CE. Retinal haemodynamics in pa­tients with early diabetes mellitus. Br .1 Ophthalmol 1996; 80: 327- 31. 70. Brown SM, Janipol LM. New concepts of regulation of retinal vessel tone. Arch Ophthalmol 1996; 114: 199- 204. 71. Sobey CG, Faraci FM, Piegors DJ, Hcislad DD. Fffccl of short-term regression of atherosclerosis on reactivity of carotid and reti­nal arteries. Stroke 1996; 27: 927- 33. 72. Jumper JM, Horton JC. Central retinal artery occlusion after ma­nipulation of the neck by a chiropractor. Am J Ophthalmol 1996; 121: 321- 2. 73. Sharma S, Naqvi A, Sharma S, Crucss AF, Brown GC, for the Retinal Emboli of the Cardiac Origin Group. Transthoracic en-chocardiographic findings in patients with acute retinal arterial obstruction. A retrospective review. Arch Ophthalmol 1996; 114: 1189- 92. 74. Fineman MS, Savino PJ, Federman JL, Eagle Jr RC. Branch retinal artery occlusion as the initial sign of giant cell arteritis. Am .! Ophthalmol 1996; 122: 428- 30. 75. The Eye Disease Case- Control Study Group. Risk factors for cen­tral retinal vein occlusion. Arch Ophthalmol 1996; 1 14: 545- 54. 76. Mitchell P, Smith W, Chang A. Prevalence and associations of retinal vein occlusion in Australia. Arch Ophthalmol 1996; 114: 1243- 7. 77. Williamson TH, Rumley A, Lowe GDO. Blood viscosity, coagu­lation, and activated protein C resistance in central retinal vein occlusion: a population- controlled study. Br .1 Ophthalmol 1996; 80: 203- 8. 78. Larsson J, Olafsdotlir E, Bauer B. Activated protein C resistance in young adults with central retinal vein occlusion. Br J Ophthalmol 1996; 80: 200- 2. 79. Yoshizumi MO, Townsend- Pico W. Essential thrombocythemia and central retinal vein occlusion with ncovascular glaucoma. Am J Ophthalmol 1996; 121: 728- 30. 80. Aritiirk N, Oge Y, Erkan D, Siillii Y, Mohajery I-'. Relation be­tween retinal vein occlusions and axial length. Br .1 Ophthalmol 1996; 80: 633- 6. 81. Glacet- Bernard A, Coscas G, Chabanel A, Zourdani A, Lelong F, Samama MM. Prognostic factors for retinal vein occlusion: a pro­spective study of 175 cases. Ophthalmology 1996; 103: 551- 60. 82. Kupersmith MJ, Vargas EM, Warren F, Bcrcnstein A. Venous obstruction as the cause of rctinal/ choroidal dysfunction associated with arteriovenous shunts in the cavernous sinus. J Neuroophthal-mol 1996; 16: 1- 6. 83. Goldberg RA, Goldey SH, Duckwilcr G, Vinuela F. Management of cavernous sinus- dural fistulas. Arch Ophthalmol 1996; 114: 707- 14. 84. Newman NJ. Optic neuropathy. Neurology 1996; 46: 315- 22. 85. Miller NR. The optic nerve. Cmr Opin Neurol 1996; 9: 5- 15. 86. Colapinto EV, Cabeen MA, Johnson LN. Optic nerve compression by a dolichoectatic internal carotid artery: case report. Neurosur­gery 1996; 39: 604- 6. 87. Golnik KC, Hund PW, Stroman GA, Stewart WC. Magnetic reso­nance imaging in patients with unexplained optic neuropathy. Oph­thalmology 1996; 103: 515- 20. ./ Neiiro- Ophthalmol, Vol. 17, No. I. 1997