Correlation of Macular Sparing and Homonymous Paracentral Scotomas With MRI Lesions in Posterior Cerebral Artery Infarction

Background: The concepts of the representation of visual field in primary visual cortex are based on studies of war wounds and correlations with brain imaging in small cohorts. Because of the difficulty of judging brain lesion extent and the small number of studied patients, there is lingering controversy as to whether the central 15° of visual field are mapped onto the posterior 25% of primary visual cortex or onto a larger area. To improve the delineation of MRI lesion extent, we have studied only patients with posterior cerebral artery (PCA) ischemic strokes. Methods: We accrued a cohort of 92 patients with PCA strokes from an electronic medical records search between 2009 and 2020 at a single tertiary care academic institution. Patients had reliable static perimetry demonstrating homonymous hemianopias and high-definition reviewable brain imaging. We divided the primary visual cortex on the MRI T1 sagittal sequence into 8 equal segments in right and left cerebral hemispheres and located lesions according to the segments they occupied. We correlated lesion locations with 3 visual field defects (VFDs): macular-sparing homonymous quadrantanopias, macular-splitting homonymous quadrantanopias, and homonymous paracentral scotomas. Results: Among 25 cases with macular sparing, 13 had lesion-sparing confined to the posterior 25% of visual cortex. Among 6 cases with homonymous paracentral scotomas, 2 had lesions confined to the posterior 25% of visual cortex. Macular-splitting quadrantanopia did not occur in any patients with lesions confined to the posterior 25% of visual cortex, but did occur in 3 patients with lesions confined to the posterior 50% of visual cortex. These phenomena would not be expected if the central 15° of visual field were mapped onto a region extending beyond the posterior 25% of visual cortex. In patients with PCA strokes that involved the retrogeniculate visual pathway proximal to visual cortex, the visual cortex lesions were often less extensive than predicted by the VFDs, perhaps because of widespread damage to axons before they reached their destination in visual cortex. Conclusions: These results support the concept that the central 15° of the visual field are represented in the posterior 25% of visual cortex. Although this study contributes a larger cohort of patients with better-defined lesion borders than in past reports, its conclusions must be tempered by the variability of patient attention during visual field testing, the subjectivity in the interpretation of the defect patterns, and the difficulty in judging MRI lesion extent even on diffusion-weighted and precontrast T1 sagittal sequences.

Original Contribution Section Editors: Clare Fraser, MD Susan Mollan, MD Correlation of Macular Sparing and Homonymous Paracentral Scotomas With MRI Lesions in Posterior Cerebral Artery Infarction Juno Cho, BS, Eric Liao, MD, Jonathan D. Trobe, MD Background: The concepts of the representation of visual field in primary visual cortex are based on studies of war wounds and correlations with brain imaging in small cohorts. Because of the difficulty of judging brain lesion extent and the small number of studied patients, there is lingering controversy as to whether the central 15° of visual field are mapped onto the posterior 25% of primary visual cortex or onto a larger area. To improve the delineation of MRI lesion extent, we have studied only patients with posterior cerebral artery (PCA) ischemic strokes. Methods: We accrued a cohort of 92 patients with PCA strokes from an electronic medical records search between 2009 and 2020 at a single tertiary care academic institution. Patients had reliable static perimetry demonstrating homonymous hemianopias and high-definition reviewable brain imaging. We divided the primary visual cortex on the MRI T1 sagittal sequence into 8 equal segments in right and left cerebral hemispheres and located lesions according to the segments they occupied. We correlated lesion locations with 3 visual field defects (VFDs): macular-sparing homonymous quadrantanopias, macular-splitting homonymous quadrantanopias, and homonymous paracentral scotomas. Results: Among 25 cases with macular sparing, 13 had lesion-sparing confined to the posterior 25% of visual cortex. Among 6 cases with homonymous paracentral scotomas, 2 had lesions confined to the posterior 25% of visual cortex. Macular-splitting quadrantanopia did not occur in any patients with lesions confined to the posterior 25% of visual cortex, but did occur in 3 patients with lesions confined to the posterior 50% of visual cortex. These phenomena would not be expected if the central 15° of visual field were mapped onto a region extending beyond the posterior 25% of visual cortex. In patients with PCA strokes that involved the retrogeniculate visual pathway proximal to visual cortex, the visual cortex lesions were often less extensive than predicted by the VFDs, perhaps because of widespread damage to axons before they reached their destination in visual cortex. Kellogg Eye Center (JC, JDT), Department of Ophthalmology and Visual Sciences, Ann Arbor, MI; Department of Radiology (Neuroradiology) (EL), Ann Arbor, MI; and Department of Neurology (JDT), University of Michigan, Ann Arbor, MI. The authors report no conflicts of interest. Address correspondence to Jonathan D. Trobe, MD, Kellogg Eye Center, 1000 Wall Street, Ann Arbor, MI 48105; E-mail: [e-mail] Cho et al: J Neuro-Ophthalmol 2022; 42: 367-371 Conclusions: These results support the concept that the central 15° of the visual field are represented in the posterior 25% of visual cortex. Although this study contributes a larger cohort of patients with better-defined lesion borders than in past reports, its conclusions must be tempered by the variability of patient attention during visual field testing, the subjectivity in the interpretation of the defect patterns, and the difficulty in judging MRI lesion extent even on diffusion-weighted and precontrast T1 sagittal sequences. Journal of Neuro-Ophthalmology 2022;42:367–371 doi: 10.1097/WNO.0000000000001603 © 2022 by North American Neuro-Ophthalmology Society T he original concepts of the representation of the visual field in primary (striate) visual cortex derive from Inouye’s studies of patients who survived wartime head trauma in the Russo-Japanese War (1) and Holmes’ studies of wounded survivors of World War I (2–4). Those investigators proposed that the central visual field mapped onto posterior visual cortex and that the peripheral field mapped onto anterior visual cortex. But without accurate methods of determining the location and extent of the wounds, they could make only crude estimations. In applying planimetry to the Holmes map in 1991, Horton and Hoyt determined that the central 15° of the visual field mapped onto the posterior 25% of visual cortex (5). Yet their own study of 3 patients and their imaging (tuberculoma, arteriovenous malformation, and infarct) led them to enlarge the representation of the central 15° of visual field to 60%–70% of posterior primary visual cortex, conforming to studies in macaque monkeys. A 1994 study of kinetic and static perimetric defects in 26 patients with occipital infarcts and mass lesions identified on computed tomography and low-resolution MRI supported the Horton–Hoyt map (6). But a 1999 correlational study of visual field defects (VFDs) and MRI-based occipital infarcts in 14 patients found more support for the Holmes map, contending that the central 15° of the visual field mapped onto the posterior 37% of visual cortex (7). 367 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution The conflicting results of these foregoing studies may be attributed to the subjective nature of visual field testing and the imprecision of imaging studies in defining lesion extent. In this study, we therefore included only patients with posterior cerebral artery (PCA) ischemic strokes, in which restricted diffusion, encephalomalacia, and laminar necrosis can define lesion extent more precisely. A consensus review by 2 authors (J.C., J.D.T.), who had no knowledge of the imaging abnormalities, defined macular sparing, macular splitting, and homonymous paracentral scotomas from the Humphrey visual field (HVF) results. We matched these 3 VFD patterns to the location of the MRI lesion with the objective of determining the proportion of posterior visual cortex that decodes signals from the central 15° of the visual field. We hypothesized that if the posterior 25% of visual cortex decodes the central 15° of visual field, then lesions that spare merely the posterior 25% of visual cortex would produce macular sparing of 15°, and lesions confined to the posterior 25% of visual cortex would produce paracentral scotomas limited to 15°. However, if a larger area of visual cortex decodes the central 15° of visual field, then lesions that spare merely the posterior 25% of visual cortex would not produce a full 15° of macular sparing, and lesions that involve the posterior 50% of visual cortex would produce paracentral scotomas limited to 15° rather than macular splitting that extends outward beyond the central 15° of the visual field. METHODS We obtained permission from the Michigan Medicine (University of Michigan) Institutional Review Board to conduct a 2009–2020 electronic medical records (Epic) search of patients with “homonymous hemianopia,” “visual fields,” and “MRI” using the university’s Electronic Medical Record Search Engine (8). We included only those patients who had MRI lesions with imaging characteristics of PCA-distribution ischemic stroke and reliable HVF homonymous hemianopias. For lesion identification, we drew on the brain MRI study performed closest to the time of the first HVF examination. The time difference between the MRI study and the HVF examination varied from 2 to 31 days, with a median of 10 days. The time difference between the stroke and the HVF varied from 1 day to 25 days, with a median of 8 days. The MRIs were of 1.5 T or 3.0 T field strength. The borders of the lesions were defined by the margins of restricted diffusion, pre-contrast T1 signal hypointensity (encephalomalacia), or T1 gyral signal hyperintensity (laminar necrosis). Where sagittal plane sequences were not available, we used 3D reconstructions. We considered the superior primary visual cortex region to lie above the calcarine fissure and the inferior primary visual cortex to lie below the calcarine fissure. We subdivided the superior and inferior primary visual cortex regions in each cerebral hemisphere into 8 segments by drawing equally spaced lines perpendicular to the calcarine fissure on the sagittal MRI. We coded the right superior bank segments 368 from rostral to caudal as R1, R2, R3, and R4, the right inferior bank segments as R5, R6, R7, and R8, the left superior bank segments as L1, L2, L3, and L4, and the left inferior bank segments as L5, L6, L7, and L8 (Fig. 1). There were 92 patients with 105 MRI hemispheric lesions (13 patients had bilateral hemispheric lesions). We divided them into 3 groups (Fig. 2): Group A: lesion limited to gray matter of primary visual cortex extending from the occipital tip to the parieto-occipital fissure; Group B: lesion involving gray matter of primary visual cortex and white matter of posterior optic radiations posteromedial to the calcar avis (atrium); Group C: lesion involving gray matter of primary visual cortex, posterior optic radiations, and anterior optic radiations. Two authors (J.C., J.D.T.) interpreted the HVF results on each patient independently and in a masked fashion. Each reader repeated the interpretations until there was consistency and consensus. The readers relied on the pattern deviation and gray scale. We included gray scale because it is commonly used by clinicians. The definitions of the VFDs were: 1. Macular sparing quadrantanopias: sparing of at least the first 2 central test points adjacent to the horizontal meridians on the pattern deviation in the affected quadrants (corresponding to 15° eccentric to fixation) in the presence of homonymous VFDs involving more peripheral regions in the affected quadrants (Fig. 3A) 2. Macular-splitting quadrantanopias: no sparing of the first 2 central test points adjacent to the horizontal meridians on the pattern deviation in the affected FIG. 1. Segmentation of primary visual cortex. On precontrast T1 sagittal MRI, we divided the visual cortex in each cerebral hemisphere into 8 segments of equal size. The superior visual cortex was coded 1–4 and the inferior visual cortex was coded 5–8. Cho et al: J Neuro-Ophthalmol 2022; 42: 367-371 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 2. The 3 groups of posterior cerebral artery strokes (diffusion-weighted images on left, corresponding apparent diffusion coefficient maps on right). A. Stroke confined to visual cortex. B. Stroke involving visual cortex and posterior optic radiations. C. Stroke involving visual cortex, and posterior and anterior optic radiations. quadrants in the presence of homonymous VFDs involving more peripheral regions in the affected quadrants (Fig. 3B) 3. Homonymous paracentral scotomas: defects involving up to the first 2 central test points (corresponding to the central 15° eccentric to fixation) without any defects affecting the visual field beyond those 2 central test points (Fig. 3C) We counted the number of macular-sparing and macular-splitting quadrantanopias in relation to lesions sparing the posterior 25% (R4, R8, L4 or L8) or the posterior 50% (R 3,4, R 7,8, L3,4 or L7,8) of the visual cortex on sagittal MRI. We then counted the number of homonymous paracentral scotomas and macular-splitting quadrantanopias in relation to lesion involvement of the posterior 25% or 50% of visual cortex. We generated the results on R software (R Core Team, Vienna, Austria, 2020) via ggplot2 package (Wickham, NY, 2014). Cho et al: J Neuro-Ophthalmol 2022; 42: 367-371 FIG. 3. Pattern deviation visual field plots of the 3 studied visual field defect patterns. A. Macular-sparing homonymous quadrantanopia. B. Macular-splitting homonymous quadrantanopia. C. Homonymous paracentral scotomas. RESULTS Macular sparing and posterior visual cortex lesion sparing (Table 1). Among the 39 cases in the entire cohort that displayed posterior visual cortex lesion sparing, there were 25 cases of macular sparing and 14 cases of macular splitting. The MRI lesion spared the posterior 25% of visual cortex in 13 cases of macular sparing (Fig. 4) and spared the posterior 50% of visual cortex in 12 cases of macular sparing. In the subgroup of cases in which the MRI lesion did not extend beyond posterior visual cortex (Group A), there were 3 cases of macular sparing associated with lesions limited to the posterior 25% of visual cortex and 6 cases associated with lesions sparing the posterior 50% of visual cortex. Macular splitting did not occur in any patients in Group A whose MRI lesions spared the posterior 25% of visual cortex. Homonymous paracentral scotomas, macular-splitting quadrantanopias, and posterior visual cortex involvement 369 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution TABLE 1. Correlation of macular-sparing and macular-splitting homonymous hemianopias with visual cortex lesion-sparing Visual Cortex Lesion Spares Only Its Posterior 25% N Groups A, B, and C Macular sparing 25 13 Macular splitting 14 12 Group A only (lesions limited to visual cortex) Macular sparing 9 3 Macular splitting 0 0 Group B only (lesions involve visual cortex and posterior optic radiations) Macular sparing 14 8 Macular splitting 9 8 Group C only (lesions involve visual cortex, posterior and anterior optic radiations) Macular sparing 2 2 Macular splitting 5 4 (Table 2). Among the 6 cases with homonymous paracentral scotomas, 2 had lesions confined to the posterior 25% of visual cortex (Fig. 5). Macular-splitting quadrantanopia did not occur in any patients with lesions confined to the posterior 25% of visual cortex, but did occur in 3 patients with lesions confined to the posterior 50% of visual cortex. DISCUSSION In this study, evidence favors the concept that the central 15° of the visual field map within the posterior 25% of visual cortex. As many as 13 cases had macular sparing associated with lesion sparing limited to the posterior 25% of visual cortex. If the central 15° of field were mapped onto an area that extended beyond that region, lesion sparing limited to only the posterior 25% of visual cortex would have been insufficient to produce macular sparing. Because strokes that displayed signal abnormalities extending in the optic radiations (Groups B and C) might have caused damage to axons before they entered visual cortex, we considered Visual Cortex Lesion Spares Its Posterior 50% 12 2 6 0 6 1 0 1 the correlation between visual cortex lesion location and macular sparing to be most trustworthy in Group A. In that subgroup, 3 cases with lesion sparing limited to the posterior 25% of visual cortex had macular sparing, a phenomenon that would not be expected if the macular representation was spread out beyond the posterior 25% of visual cortex. In the 6 remaining cases that had macular sparing in association with lesion sparing that extended into the posterior 50% of visual cortex, we posit that the macular sparing might have resulted from the component of the lesion that spared the posterior 25% of visual cortex. Although there were only 6 cases with homonymous paracentral scotomas, 2 of them were associated with lesions limited to the posterior 25% of visual cortex, further supporting the hypothesis that the central 15° of visual field are represented in that small region of visual cortex. Finally, macular-splitting quadrantanopia occurred in 3 patients with lesions confined to the posterior 50% of visual cortex, a phenomenon that would not be expected if the central 15% of the visual field were represented in such a large extent of posterior visual cortex. TABLE 2. Correlation of homonymous paracentral scotomas and macular-splitting quadrantanopias with posterior visual cortex lesion involvement N Visual Cortex Lesion Confined to Its Posterior 25% Groups A, B, and C Homonymous paracentral scotomas 6 2 Macular-splitting quadrantanopias 3 0 Group A only (lesions limited to visual cortex) Homonymous paracentral scotomas 2 2 Macular-splitting quadrantanopias 1 0 Group B only (lesions involve visual cortex and posterior optic radiations) Homonymous paracentral scotomas 2 0 Macular-splitting quadrantanopias 2 0 Group C only (lesions involve visual cortex, and posterior and anterior optic radiations) Homonymous paracentral scotomas 2 0 Macular-splitting quadrantanopias 0 0 370 Visual Cortex Lesion Confined to Its Posterior 50% 4 3 0 1 2 2 2 0 Cho et al: J Neuro-Ophthalmol 2022; 42: 367-371 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 4. Pattern deviation visual field plots display macular sparing in left superior and inferior homonymous quadrants (left). Precontrast T1 sagittal MRI shows gyral high signal affecting anterior 75% of right superior and inferior visual cortex, sparing its posterior 25% (right). Dotted red line traces course of calcarine fissure. FIG. 5. Pattern deviation visual field plots display homonymous paracentral scotomas involving right superior and inferior quadrants (left). Precontrast T1 sagittal MRI shows encephalomalacia confined to posterior 25% of left superior and inferior visual cortex (right). Dotted red line traces course of calcarine fissure. Macular-splitting quadrantanopia occurred in 14 cases, 12 of which had lesion sparing limited to the posterior 25% of visual cortex. All 14 of these cases had lesions extending into the posterior optic radiations (Group B) or the anterior optic radiations (Group C). Therefore, we propose that this macular splitting occurred because axons were damaged before they reached visual cortex. Thus, in proximal infarcts, the size of the lesion in visual cortex would underestimate the extent of the VFD. Support for that hypothesis comes from the fact that this phenomenon occurred exclusively in Groups B and C. Several study limitations temper our conclusions. First, the number of patients with macular sparing and paracentral scotomas was relatively small. Second, although our interpretations of the VFDs were based on several iterations until consensus was reached, 1 reader was trained by the other reader. Moreover, we were obliged to rely heavily on thresholds at a small number of central threshold points that may have been markedly affected by whether patients fixated properly. Third, although definition of the MRI lesions was mostly based on relatively discrete borders generated by restricted diffusion, encephalomalacia, and gyral necrosis, those borders were sometimes indistinct. The circuitous path of the calcarine fissure caused segments to disappear from view, making it sometimes difficult to locate the lesion in relation to the fissure. Despite these limitations, we believe this study to be more precise than its predecessors in providing evidence of where visual pathway axons land on visual cortex. Cho et al: J Neuro-Ophthalmol 2022; 42: 367-371 STATEMENT OF AUTHORSHIP Conception and design: J. Cho, E. Liao, J. D. Trobe; Acquisition of data: J. Cho, Eric 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. REFERENCES 1. Inouye T. Die Sehstorungen bei Schussverletzungen der kortikalen Sehsphare. Leipzig, Germany: W Engelmann; 1909. 2. 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. 3. Holmes G. Disturbances of vision by cerebral lesions. Br J Ophthalmol. 1917;2:353–384. 4. Holmes G. The organization of the visual cortex in man. Proc R Soc Lond Ser B (Biol).1945;132:348–361. 5. 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. 6. 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. 7. Wong AM, Sharpe JA. Representation of the visual field in the human occipital cortex: a magnetic resonance imaging and perimetric correlation. Arch Ophthalmol. 1999;117:208–217. 8. 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. 371 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.