Natural History of Stroke-Related Homonymous Hemianopsia

Abstract & Commentary

By Marc Dinkin, MD, Assistant Professor of Ophthalmology, Weill Cornell Medical College. Dr. Dinkin reports no financial relationships relevant to this field of study.

Synopsis: Serial examination of Humphrey visual fields in patients with infarct-related hemianopsia shows spontaneous recovery, predominantly in lower quadrants and in the periphery.

Source: Çelebisoy M, et al. Recovery of visual-field defects after occipital lobe infarction: A perimetric study. J Neurol Neurosurg Psychiatry 2011;82:695-702.

Nearly one-third of stroke victims suffer from hom- onymous hemianopsias related to damage to the optic radiations or occipital cortex. These defects may result in a significant change in quality of life and independence. Following such strokes, patients typically want to know how much improvement there will be and when it will occur. Many are referred to intensive rehabilitation programs aimed at visual restoration, but differentiating therapy-related improvement from spontaneous recovery has been difficult. In this study, neurologists prospectively studied the natural history of homonymous hemianopsia in 32 patients following ischemic stroke, comparing fields acquired within 2 weeks of their stroke to those completed at 6 months. Similar studies had been performed before using either confrontational or manual (Goldmann) visual fields, but this was the first to utilize Humphrey automated perimetry, which allows for a more rigorous statistical analysis. A mean visual sensitivity score (MVS) was generated for central, mid-peripheral, and peripheral sectors in each quadrant and compared in both the affected and unaffected side. The effect of various locations for the infarct (optic radiation, occipital pole, occipital convexity, cuneus) was analyzed.

Although prior investigations had demonstrated some degree of improvement in > 50% of patients, this study clarified that this applied to defects in the lower quadrants and predominantly in the central and paracentral zones. All zones showed an improvement in the median or maximum MVS for the cohort. Since it is known that visual field improvement occurs by virtue of practice, a comparison was made with the unaffected side where the majority of patients also showed improvement. However, the improvement in the affected hemifields was significantly greater than that of the unaffected fields in the lower zones, suggesting a true improvement in these regions. Interestingly, those infarcts that did not include the striate cortex were associated with a higher chance of recovery, which was most prominent in the paracentral and peripheral zones, most significantly in the lower quadrants. Not surprisingly, patients with smaller lesions containing only one or two of the designated areas showed a greater degree of improvement than those with larger infarcts.


This study by Çelebisoy et al makes a significant contribution to our knowledge of the spontaneous recovery that occurs after occipital lobe infarcts. The authors confirmed prior findings of (1) greater improvement in the lower quadrants, and (2) more prominent restoration of the fields in the periphery. Their findings stand up to the more rigorous statistical analysis afforded by Humphrey automated perimetry. The gross measurement of percentage change in Messing's paper1 is replaced with comparisons of quantitative mean deviations in each field region.

It is hypothesized that the greater degree of improvement in the lower quadrants results from supplementary perfusion by the parieto-occipital artery to the superior striate cortex, although no explanation is given for why the presence of redundant perfusion by the posterior temporal artery to the inferior cortex would not have the same effect on superior field recovery. The authors echo Messing's theory that recovery of equal areas of peripheral cortex result in greater visual field recovery than that of central cortex given the greater receptive fields in the periphery.

There were several limitations to this study. First, serial examinations at intermediate time points were not performed, so the timing of the visual field improvement within the 6 months is not elucidated and recovery beyond 6 months cannot be ruled out. Unlike the Messing study, they did not include lateral geniculate lesions or look at hemorrhagic stroke. The schematic divisions of the visual cortex based on Messing's article seem somewhat arbitrary with the medial portion of the occipital lobe designated as striate cortex and the occipital pole delineated as if it is extra-striate. Moreover, no details are offered as to how stroke location was determined based on this scheme or if radiologists were involved. There was no review of the degree of edema, the mechanism of infarction (thrombotic vs. embolic), or the presence or absence of treatment, and their potential relationship with recovery. Finally, since the authors speculated that supplementary blood flow might play a role in recovery, an analysis of the MR angiograms of these patients comparing the degree of redundant circulation with recovery would have helped put that theory to the test.

Despite these limitations, this study remains a strong addition to the relatively small body of literature on the natural history of homonymous visual field defects. Most importantly, it confirms that there is a potential for significant recovery in patients following occipital lobe stroke, giving hope to stroke victims suffering from great disability. It also serves as a reminder that any visual restoration therapy needs to be measured against this spontaneous recovery, so that patients are not made to erroneously attribute such improvement to a timely and often costly rehabilitation program.


1. Messing B, Ganshirt H. Follow-up of visual field defects with vascular damage of the geniculostriate visual pathway. J Neuroophthalmol 1987;7:231-242.