Visual Field Defect Patterns Associated With Lesions of the Retrochiasmal Visual Pathway

Background: Perimetry is widely used in the localization of retrochiasmal visual pathway lesions. Although macular sparing, homonymous paracentral scotomas, and quadrantanopias are regarded as features of posterior retrochiasmal visual pathway lesions, incongruous hemianopia is regarded as a hallmark of anterior lesions. Recent studies have questioned the specificity of these defect patterns. Methods: Retrospective record review conducted in a single, academic, medical center using an electronic search engine with the terms ''homonymous hemianopia,' 'optic tract,' 'temporal lobectomy,' 'visual field defect,' and 'MRI.' Patients were included if they had reliable, automated, static visual fields, high-quality reviewable MRI scans, and pertinent lesions. MRI lesions were assigned to 1 of 6 retrochiasmal visual pathway segments by the study neuroradiologist. Two study authors independently reviewed the visual fields and designated 10 different defect patterns. Results: From an original cohort of 256 cases, only 83 had MRI-defined lesions that were limited to particular retrochiasmal segments and had visual field defect patterns that allegedly permitted localization to those particular segments. The 5 contralateral nerve fiber bundle defects were exclusive to optic tract tumors with rostral extension. Pie-in-the-sky defects were exclusive to Meyer loop lesions. Among 22 fields with macular sparing, 86% arose from the visual cortex or posterior optic radiations. Among 31 fields with homonymous quadrantanopias, 77% arose from Meyer loop, visual cortex, or posterior optic radiations. Among 13 fields with homonymous paracentral scotomas, 69% arose from visual cortex or posterior optic radiations. Optic tract lesions accounted for 70% of incongruous hemianopias but that pattern occurred uncommonly. Conclusion: In correlating discrete MRI-defined retrochiasmal lesions with visual field defect patterns identified on static perimetry, this study showed that macular sparing, homonymous paracentral scotomas, and quadrantanopias localized to the visual cortex and posterior optic radiations segments but not exclusively. It has differed from an earlier study in showing that incongruous hemianopias occur predominantly from optic tract lesions.

Original Contribution Section Editors: Clare Fraser, MD Susan Mollan, MD Visual Field Defect Patterns Associated With Lesions of the Retrochiasmal Visual Pathway Juno Cho, MA, Eric Liao, MD, Jonathan D. Trobe, MD Background: Perimetry is widely used in the localization of retrochiasmal visual pathway lesions. Although macular sparing, homonymous paracentral scotomas, and quadrantanopias are regarded as features of posterior retrochiasmal visual pathway lesions, incongruous hemianopia is regarded as a hallmark of anterior lesions. Recent studies have questioned the specificity of these defect patterns. Methods: Retrospective record review conducted in a single, academic, medical center using an electronic search engine with the terms ““homonymous hemianopia,” “optic tract,” “temporal lobectomy,” “visual field defect,” and “MRI.” Patients were included if they had reliable, automated, static visual fields, high-quality reviewable MRI scans, and pertinent lesions. MRI lesions were assigned to 1 of 6 retrochiasmal visual pathway segments by the study neuroradiologist. Two study authors independently reviewed the visual fields and designated 10 different defect patterns. Results: From an original cohort of 256 cases, only 83 had MRI-defined lesions that were limited to particular retrochiasmal segments and had visual field defect patterns that allegedly permitted localization to those particular segments. The 5 contralateral nerve fiber bundle defects were exclusive to optic tract tumors with rostral extension. Pie-inthe-sky defects were exclusive to Meyer loop lesions. Among 22 fields with macular sparing, 86% arose from the visual cortex or posterior optic radiations. Among 31 fields with homonymous quadrantanopias, 77% arose from Meyer loop, visual cortex, or posterior optic radiations. Among 13 fields with homonymous paracentral scotomas, 69% arose from visual cortex or posterior optic radiations. Optic tract lesions accounted for 70% of incongruous hemianopias but that pattern occurred uncommonly. Kellogg Eye Center, Departments of Ophthalmology and Visual Sciences (JC, JDT), Radiology (Neuroradiology) (EL), and Neurology (JDT), University of Michigan, Ann Arbor, Michigan. Supported by National Center for Advancing Translational Sciences (Bethesda, MD, TL1TR002242, J.C.). The authors report no conflicts of interest. The funding organizations had no role in the design and conduct of the study, collection, management, analysis and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Address correspondence to Jonathan D. Trobe, MD, Kellogg Eye Center, Department of Ophthalmology and Visual Sciences, University of Michigan, 1000 Wall Street, Ann Arbor, MI 48105; E-mail: [e-mail] Cho et al: J Neuro-Ophthalmol 2022; 42: 353-359 Conclusion: In correlating discrete MRI-defined retrochiasmal lesions with visual field defect patterns identified on static perimetry, this study showed that macular sparing, homonymous paracentral scotomas, and quadrantanopias localized to the visual cortex and posterior optic radiations segments but not exclusively. It has differed from an earlier study in showing that incongruous hemianopias occur predominantly from optic tract lesions. Journal of Neuro-Ophthalmology 2022;42:353–359 doi: 10.1097/WNO.0000000000001601 © 2022 by North American Neuro-Ophthalmology Society T he measurement of patient-identified achromatic visual thresholds in the field of vision is a time-honored diagnostic tool in localizing visual pathway lesions (1). As such, it is valuable in guiding and confirming neuroradiologic investigation. The ability of visual field examination to localize lesions is based on the fact that certain defect patterns have been linked by imaging, biopsy, and necropsy to lesions in specific segments of the visual pathway (1–7). For example, homonymous quadrantanopias, homonymous hemianopias with macular sparing, homonymous paracentral scotomas, temporal crescent-sparing homonymous hemianopias, and unilateral temporal crescent defects have been associated with lesions of the primary visual cortex (1,3–5). Wedgeshaped homonymous hemianopias in the superior visual field (“pie-in-the-sky” defects) have been considered hallmarks of lesions of Meyer loop in the anterior temporal lobe (1). Incongruous homonymous hemianopias have been linked to lesions of the optic tract (6), whereas congruous homonymous hemianopias have been linked to lesions of the retrogeniculate segment (1). However, in a study of 904 patients with retrochiasmal lesions using static or kinetic visual field testing and CT or MRI, the authors stated that the visual field defect patterns were not as location specific as previously supposed (8–10). We have investigated the correlation of homonymous hemianopic defects and brain lesions using different 353 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution methods. Unlike earlier studies (8–10), this investigation studied patients whose visual field defects were identified exclusively on automated visual field (HVF) testing and whose lesions were confined to a particular retrochiasmal segment on high-definition MRI available for our review. METHODS We obtained permission from the Michigan Medicine (University of Michigan) Institutional Review Board to conduct a 1990–2020 electronic medical records (Epic) search of patients with “homonymous hemianopia,” “optic tract,” “temporal lobectomy,” “visual field defect,” and “MRI” using the Electronic Medical Record Search Engine (EMERSE) (11). We included patients who had a homonymous hemianopia identified on a reliable HVF examination performed within 2 years (90% within 186 days) of the initial diagnosis at Michigan Medicine and who had reviewable high-quality brain MRI study conducted within that period. From the medical records, we excerpted patient age, sex, date of initial diagnosis at Michigan Medicine, cause of the lesion, and pertinent ophthalmologic and neurologic data from the initial examination. For lesion identification, we used the brain MRI study performed closest to the time of the initial diagnosis. The anatomical location of the pertinent lesion was determined entirely by the neuroradiologist author (E.L.), who reviewed all brain MRIs with the clinical information provided on the imaging request but without knowledge of the HVF results. Based on a review of a composite of pulse sequences, the lesion was assigned to 1 of the following 6 segments of the retrochiasmal visual pathway (Fig. 1): 1. Optic tract with rostral extension: lesions based in the optic tract on 1 side but with major rostral extension to involve the optic chiasm and/or optic nerves (Fig. 1A) 2. Optic tract confined: lesions limited to the optic tract on 1 side (Fig. 1B) 3. Anterior optic radiations: lesions involving the optic radiations from their root to the posterior border of the atrium of the lateral ventricle without apparent extension into Meyer loop (Fig. 1C) 4. Meyer loop: lesions confined to the anterior temporal lobe on 1 side without extension into the region of the lateral geniculate body or root of the optic radiations; all lesions resulted from temporal lobectomy (Fig. 1D) 5. Visual cortex and posterior optic radiations: lesions involving primary visual cortex with extension into posterior optic radiations, the anterior border of which was the atrium of the lateral ventricle (Fig. 1E) 6. Visual cortex confined: lesions limited to the region of primary visual cortex (Fig. 1F) The HVFs were performed with the 24-2 threshold test protocol. The defect patterns were identified independently FIG. 1. The 6 anatomical segments of the retrochiasmal visual pathway and the corresponding representative axial MRI scans. 354 Cho et al: J Neuro-Ophthalmol 2022; 42: 353-359 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution and in masked fashion by 2 authors (J.C. and J.D.T.) without any clinical information or knowledge of the location of the pertinent MRI lesion. The interpretations were repeated independently by both readers until there was consistency and consensus. In their interpretations, the authors relied on the gray scale in combination with the pattern deviation. The gray scale was included because it is commonly relied on by clinicians. The defect patterns were as follows (Fig. 2): 1. Uninterpretable: no discernible localizing defect pattern (Fig. 2A). 2. Complete homonymous hemianopia: maximally high thresholds involving all test points in both quadrants of the affected hemifields in both eyes (Fig. 2B). 3. Quadrantanopia: defects limited to single homonymous quadrants in both eyes with borders aligned to the vertical and horizontal meridians (Fig. 2C). FIG. 2. Representative automated visual field gray scale plots of the 10 visual field defect patterns identified in the original patient cohort. Only 7 of these patterns had localizing features that permitted correlation with MRI lesions. Cho et al: J Neuro-Ophthalmol 2022; 42: 353-359 355 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution 4. Macular sparing: relative sparing of at least the first 2 central test points adjacent to the horizontal meridians in the affected hemifields of both eyes without sparing of the 2 most peripheral test points (Fig. 2D). 5. Homonymous paracentral scotomas: relative sparing of at least the 2 most peripheral test points adjacent to the horizontal meridian in the affected homonymous hemifields of both eyes without sparing of the 2 most central test points (Fig. 2E). 6. Superior homonymous wedge (“pie-in-the-sky”) defects: defects in the superior hemifields of both eyes with relative sparing of at least the first 2 central test points and 2 most peripheral test points adjacent to the horizontal meridians in the superior hemifields (Fig. 2F). 7. Inferior homonymous wedge (“pie-on-the-floor”) defects: defects in the inferior hemifields of both eyes with relative sparing of at least the first 2 central test points and 2 most peripheral test points adjacent to the horizontal meridians in the inferior hemifields (Fig. 2G). 8. Homonymous hemianopia with contralateral nerve fiber bundle defect: arcuate defect occupying the superior and/or inferior field in the eye contralateral to the side of the hemianopia (Fig. 2H). 9. Incomplete homonymous hemianopia without quadrantanopia, macular sparing, homonymous paracentral scotomas, pie-in-the-sky, or pie-on-the-floor: homonymous hemianopic defects that lacked these features (Fig. 2I). 10. Incongruous homonymous hemianopia: incomplete homonymous hemianopic defects that appeared substantially unequal in size and depth in the 2 eyes (Fig. 2J). There were no cases that showed sectoranopia, unilateral temporal crescent defects, or temporal crescent sparing. RESULTS From EMERSE (11), we identified 1,507 cases whose records contained our keywords. Only 83 visual fields belonged to patients with localizable visual field defects and MRI-defined lesions within the retrochiasmal pathway that could be securely assigned to one of the predetermined segments of the pathway. Most of the excluded cases were large lesions caused by tumors or proximal posterior cerebral artery ischemic strokes. Seven of the 10 identified visual field defect patterns were suitable for correlation with lesion location (Table 1 and Fig. 3). Quadrantanopia Of the 31 cases with this defect pattern, 24 (77%) arose from the visual cortex, posterior optic radiations, or Meyer loop. The remaining 7 (23%) derived from the optic tract or the anterior optic radiations (Fig. 3A). Macular Sparing Of the 22 cases with this defect pattern, 19 (86%) arose from the visual cortex or posterior optic radiations. The remaining 3 (14%) arose from lesions of the optic tract or anterior optic radiations (Fig. 3B). Homonymous Paracentral Scotomas Of the 13 cases with this defect pattern, 9 (69%) arose from lesions in the visual cortex or posterior optic radiations. The remaining 4 (31%) arose from optic tract or anterior optic radiations. The defect patterns generated TABLE 1. Correlation of 6 retrochiasmal visual pathway segments with 7 visual field defect patterns Location of MRI Lesions Optic tract with rostral extension (n = 17) Optic tract confined (n = 14) Anterior optic radiations (n = 12) Meyer loop (n = 9) Visual cortex and posterior optic radiations (n = 40) Visual cortex confined (n = 30) 356 Homonymous Inferior Superior Homonymous Homonymous Hemianopia With Homonymous wedge-Shaped Wedge-Shaped Contralateral Nerve Fiber Macular Paracentral (“Pie-in-the-Sky”) (“Pie-on-theFloor”) Defects Bundle Defect Incongruity Defects Quadrantanopia Sparing Scotomas 2 0 0 0 0 5 5 2 1 1 0 1 0 2 3 2 3 1 0 0 1 4 10 0 8 0 4 5 0 0 2 0 0 0 1 10 11 5 0 2 0 1 Cho et al: J Neuro-Ophthalmol 2022; 42: 353-359 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 3. Bar graphs showing the distribution of the 7 localizing visual field defect patterns across 6 anatomical segments of the retrochiasmal visual pathway. by lesions of the optic tract and anterior optic radiations had no features that distinguished them from the defects generated by lesions of the posterior optic radiations and visual cortex (Fig. 3C). the anterior optic radiations that appeared to extend into Meyer loop (Fig. 3D). Superior Homonymous Wedge-Shaped (“Pie-inthe-Sky”) Defects Among the 5 cases with this defect pattern, 4 (80%) arose from visual cortex or posterior optic radiations, and 1 (20%) arose from a lesion confined to the optic tract (Fig. 3E). Among the 6 cases with this defect pattern, 5 (83%) arose from lesions in Meyer loop and 1 (17%) from a lesion in Cho et al: J Neuro-Ophthalmol 2022; 42: 353-359 Inferior Homonymous Wedge-Shaped (“Pie-onthe-Floor”) Defects 357 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution Homonymous Hemianopia With Contralateral Nerve Fiber Bundle Defect All 5 cases with this defect pattern derived from lesions in the optic tract extended rostrally to involve the ipsilateral optic nerve (Fig. 3F). Incongruity Among the 107 incomplete hemianopias, only 10 (9%) were incongruous. Of the 10 incongruous visual field cases, 7 (70%) arose from the optic tract and 3 (30%) from the visual cortex, posterior optic radiations, or anterior optic radiations (Fig. 3G). CONCLUSIONS This study of retrochiasmal visual pathway lesions and their associated visual field defects has validated the classical concepts of perimetric localization. The pattern of a contralateral nerve fiber bundle defect in combination with a homonymous hemianopia was found exclusively in tumorous optic tract lesions that extended rostrally into 1 optic nerve. It did not occur in any of the nontumorous lesions confined to the optic tract or in any other segment of the retrochiasmal visual pathway. Buried within the hemianopic defects, the nerve fiber bundle defects were often hard to detect. As a reflection of damage to the ipsilateral optic nerve, this pattern is a useful marker of tumor in that location. The combination of optic tract and ipsilateral optic nerve dysfunction with tumors in this location has been previously recognized by an afferent pupil defect and optic disc pallor contralateral to the hemianopia (1,6,12), but the localizing value of the nerve fiber bundle defect has not been emphasized. Superior, homonymous, wedge-shaped (“pie-in-thesky”) defects occurred exclusively with Meyer loop lesions. This pattern was seen in 5 of 9 Meyer loop lesions and in 1 lesion of the anterior optic radiations that appeared to extend into Meyer loop. Notably, 4 temporal lobectomies produced full quadrantanopias rather than pie-in-the-sky defects. These large visual field defects, which had inferior borders extending down to the horizontal meridian, were associated with lobectomies whose borders reached posteriorly beyond 5 cm from the anterior temporal lobe tip. In those cases, the resection cavity presumably extended backward into the anterior optic radiations. In previously reported series of temporal lobectomies, there has been considerable variation in the size of the excision needed to generate a visual field defect (1). An excision of less than 4 cm has not caused a visual field defect, whereas an excision of more than 8 cm has caused a complete homonymous hemianopia (1). Although our cases were too few to permit firm conclusions, we caution that excisions extending beyond 5 cm could generate full quadrantanopias that might preclude safe driving. 358 Inferior, homonymous, wedge-shaped (“pie-on-thefloor”) defects arose from 4 lesions located in visual cortex and 1 lesion located in the optic tract. These rarely described defects might well be overlooked or misinterpreted because they are small and situated far from fixation (Fig. 2G). Macular sparing, regarded as a feature of visual cortex lesions that do not damage its most posterior portion (1,3), was relatively specific to lesions confined to visual cortex but also occurred with lesions that extended rostrally into posterior optic radiations, suggesting that this defect pattern may occur even with lesions that appear to damage juxtacortical axons. By contrast, macular sparing appeared in only 6% of lesions situated more anteriorly in the retrochiasmal pathway, implying that macular sparing is a characteristic of far posterior retrochiasmal visual pathway lesions rather than merely the result of poor fixation (1). Homonymous paracentral scotomas, traditionally associated with visual cortex lesions that spare its anterior portion (1,3,4,7,13), were considerably less common than macular sparing. Although they were mostly associated with lesions in the posterior retrochiasmal pathway, a minority occurred with lesions of the anterior optic radiations, a phenomenon that is anatomically unexplained (1). Quadrantanopias, defined in this study as having substantial borders aligned to both vertical and horizontal meridians, are considered characteristic of visual cortex lesions (1,3) because the anatomical quadrants of visual cortex are often selectively damaged by ischemic stroke or other lesions. In this study, the posterior retrochiasmal pathway did account for most of the cases with this defect pattern. A minority of quadrantanopias arose from lesions based in the anterior optic radiations and optic tract, an unexpected occurrence that has no anatomical explanation. Incongruity occurred in a minority of our cases, mostly with lesions of the optic tract, as classically described (1,14). Our results differ somewhat from those of the major comparison study, which surveyed a larger cohort of retrochiasmal lesions but included cases identified on a mixture of CT and MRI and with visual field defects found on kinetic as well as static perimetry (8,9). In that study, the authors subdivided the retrochiasmal pathway into optic tract, optic radiations, and visual cortex and did not distinguish between lesions of the anterior and posterior optic radiations. In that study, macular sparing, homonymous paracentral scotomas, and quadrantanopias occurred after lesions in the optic tract more often than in our study. It also concluded that incongruity occurred in a substantial proportion of posterior retrochiasmal lesions, a finding at variance with our results. We submit that determining if an incomplete homonymous hemianopia is congruous or incongruous is highly subjective, as evidenced by the fact the 2 authors who interpreted the visual fields in our study had considerable difficulty reaching a consensus on this point. This study is limited by the relatively small size of the cohort. But to draw a meaningful correlation between visual Cho et al: J Neuro-Ophthalmol 2022; 42: 353-359 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution field defect patterns and their origin within the retrochiasmal pathway, we felt obliged to exclude a preponderance of cases identified in our search. Only a minority were excluded because of nonlocalizing visual field defect patterns. Most cases were unsuitable because they had lesions that spanned more than 1 visual pathway segment. Even among the lesions confined to 1 visual pathway segment, estimating their borders was challenging. Diffusion-weighted imaging and T1-weighted sequences did a reasonable job in defining lesion borders, but sometimes, those borders were more extensive on T2-weighted sequences, forcing us to choose which sequences were relevant. Despite these limitations, we are confident that the 83 included cases represent a rigorously curated sample. As such, this study affirms a strong tendency for certain visual field defect patterns to associate with particular segments of the retrochiasmal pathway, justifying the use of perimetry in localizing lesions and in serving as an aid to radiologists, especially when the imaging features are subtle. STATEMENT OF AUTHORSHIP Conception and design: J. Cho, E. Liao, J. D. Trobe; Acquisition of data: J. Cho, E. Liao, J. D. Trobe; Analysis and interpretation of data: J. Cho, E. Liao, J. D. Trobe. Drafting the manuscript: J. Cho, E. Liao, J. D. Trobe; Revising the manuscript for intellectual content: J. Cho, E. Liao, J. D. Trobe. Final approval of the completed manuscript: J. Cho, E. Liao, J. D. Trobe. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. REFERENCES 1. Miller NR, Newman NJ. Chapter 8. Topical diagnosis of lesions in the visual sensory pathway. In: Miller NR, Newman NJ, eds. Cho et al: J Neuro-Ophthalmol 2022; 42: 353-359 13. 14. Walsh & Hoyt’s Clinical Neuro-Ophthalmology. 5th edition. Volume 1. Baltimore, MD: Williams & Wilkins, 2005:237–386. Horton JC, Hoyt WF. Quadrantic visual field defects. A hallmark of lesions in extrastriate (V2/V3) cortex. Brain. 1991;114(pt 4):1703–1718. Horton JC, Hoyt WF. The representation of the visual field in human striate cortex. A revision of the classic Holmes map. Arch Ophthalmol. 1991;109:816–824. McFadzean R, Brosnahan D, Hadley D, Mutlukan E. Representation of the visual field in the occipital striate cortex. Br J Ophthalmol. 1994;78:185–190. Holmes G, Lister WT. Disturbances of vision from cerebral lesions, with special reference to the cortical representation of the macula. Brain. 1916;39:34–73. Savino PJ, Paris M, Schatz NJ, Orr LS, Corbett JJ. Optic tract syndrome. A review of 21 patients. Arch Ophthalmol. 1978;96:656–663. Spalding JM. Wounds of the visual pathway. Part II. The striate cortex. J Neurol Neurosurg Psychiatry. 1952;15:169– 183. Zhang X, Kedar S, Lynn MJ, Newman NJ, Biousse V. Homonymous hemianopias: clinical-anatomic correlations in 904 cases. Neurology. 2006;66:906–910. Zhang X, Kedar S, Lynn MJ, Newman NJ, Biousse V. Homonymous hemianopia in stroke. J Neuroophthalmol. 2006;26:180–183. Kedar S, Zhang X, Lynn MJ, Newman NJ, Biousse V. Congruency in homonymous hemianopia. Am J Ophthalmol. 2007;143:772–780. Hanauer DA, Mei Q, Law J, Khanna R, Zheng K. Supporting information retrieval from electronic health records: a report of University of Michigan’s nine-year experience in developing and using the Electronic Medical Record Search Engine (EMERSE). J Biomed Inform. 2015;55(suppl C):290–300. Anderson DR, Trobe JD, Hood TW, Gebarski SS. Optic tract injury after anterior temporal lobectomy. Ophthalmology. 1989;96:1065–1070. Kölmel HW. Homonymous paracentral scotomas. J Neurol. 1987;235:22–25. Smith JL. Homonymous hemianopia. A review of one hundred cases. Am J Ophthalmol. 1962;54:616–623. 359 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.