Abstract
Objective: To describe the characteristics of spontaneous recovery of homonymous hemianopia (HH).
Methods: The authors reviewed medical records of all patients with HH confirmed by formal visual field testing and seen in follow-up in their service between 1989 and 2004. Clinical characteristics, causes, neuroradiologic definition of lesion location, final outcome, and evolution of the visual field defects were recorded. The associations among final visual field defect outcome, time from injury, and clinical features were analyzed.
Results: A total of 254 patients with 263 HH were included in this study. Spontaneous visual field defect recovery was observed in 101 HH (38.4%). The likelihood of spontaneous recovery decreased with increasing time from injury to initial visual field testing (p = 0.0003). The probability of improvement was related to the time since injury (p = 0.0003) with a 50 to 60% chance of improvement for cases tested within 1 month after injury that decreased to about 20% for cases tested at 6 months after surgery. No other factor was found to correlate with the final outcome of the visual field defects. Improvement after 6 months from injury was mild and usually related to improvement of the underlying disease.
Conclusion: Spontaneous improvement of homonymous hemianopia is seen in at least 50% of patients first seen within 1 month of injury. In most cases, the improvement occurs within the first 3 months from injury. Spontaneous improvement after 6 months postinjury should be interpreted with caution as it is most likely related to improvement of the underlying disease or to improvement in the patient’s ability to perform visual field testing reliably.
Homonymous hemianopia (HH) is a common finding in patients with cerebral injuries.1 Some recent studies have suggested that visual restoration therapies may improve HH recovery.2–4 However, little is known about spontaneous recovery of HH.5–14 Indeed few reports have detailed the natural course of HH (see table E-1 on the Neurology Web site at www.neurology.org). Only two of the relatively large clinical studies of HH8,14 mentioned the final outcome of visual field defects; however they did not address the time course of recovery and potential related factors. Two prospective studies12,13 suggested that complete HHs mostly recover within the first 2 weeks after a stroke; however, one of these studies12 only evaluated patients with confrontation visual field testing. Finally, another study11 reported that the functional restoration was completed within 6 months and that the extent of the improvement was related to the size and location of lesion. However, these three studies11–13 were limited to stroke-related HH and the duration of follow-up varied (table E-1). Some authors believe that patients still have a chance of recovery more than 6 months after a stroke and that spontaneous recovery can no longer be expected after 1 year.15 It is likely that better knowledge of the natural history of HH would help determine the optimal “treatment window” for visual rehabilitation.1–4,15 In this study, we describe the temporal course of HH.
Methods.
Selection of patients.
As detailed in another report in this issue,16 medical records of all patients with HH seen in the Neuro-Ophthalmology Unit at Emory University between August 1989 and June 2004 were retrospectively reviewed. The study was approved by the Emory University Institutional Review Board. Only patients with confirmed HH by formal visual field testing (Goldmann [GVF] or Humphrey visual fields [HVF] tests), detailed clinical information, results of head CT or brain MRI, and at least one follow-up at our institution were included. Patients who had only confrontation visual field testing were not included. Demographic characteristics, clinical features, characteristics of visual field defects, causes of HH, neuroradiologic definition of lesion location, and associated neurologic deficits were recorded. Cases with bilateral HH were recorded twice so that each HH (right and left) could be analyzed separately. All GVF tests were performed by the same experienced technician.
Visual field testing.
All visual field tests were interpreted by two investigators (X.Z. and S.K.). In case of discrepancy, the visual fields were reviewed by two experienced neuro-ophthalmologists (V.B. and N.J.N.). Humphrey visual fields were performed on the 24-2 or 30-2 full threshold program until 2000 and on the SITA-fast or SITA-standard programs since 2001. All GVF tests were performed in a standardized way by the same experienced technician.16 Homonymous hemianopias were separated into complete and incomplete HH as detailed in another report.16 Incomplete HH included homonymous quadrantanopia, partial HH, HH with macular sparing, homonymous scotomatous defects, homonymous sectoranopia, and unilateral loss of temporal crescent. The location of brain lesion was determined based on the head CT or brain MRI report.16
Time from onset of brain injury to the initial visit in our service (corresponding to the first formal visual field testing), duration of follow-up, final neurologic outcome, and changes in visual field defects at <1, 3, 6, 9, 12, 18, 24, and greater than 24 months from injury were recorded when available. Changes of visual field defects were classified as normal, better, stable, or worse. Better or worse were defined by differences between two consecutive visual field tests of more than 10 degrees horizontally or 15 degrees vertically. Total and pattern deviation were used for HVF and similar isopters were used for comparison of GVF tests. For analysis, normal and better visual field tests were grouped into improved, while stable and worse visual field tests were grouped into nonimproved HH. Changes in visual field defects at the last available follow-up visit were established by comparing the initial with the last visual field tests.
Statistical methods.
The percent of cases that improved was compared among intervals of time from injury to initial visual field test using a χ2 test. A Cochran-Armitage test for trend was done to determine if there was a systematic change in the percent of cases that improved with increasing time from injury to initial visual field test. Tukey’s multiple comparison procedure was used to compare the percent of cases that improved between all possible pairs of time intervals. The relationship between the probability of improvement and the time from injury to initial visual field test was assessed using logistic regression. An independent groups t test or Wilcoxon rank sum test was used to compare the mean or median of continuous variables between groups of patients that either improved or did not improve; percentages for categorical variables were compared between these two groups using a χ2 test.
Results.
Clinical features and visual field defects at baseline.
Among the 880 consecutive patients with HH seen in our unit between 1989 and 2004, 254 were included in this study. Twenty-eight patients were excluded because of incomplete clinical or radiologic data; 40 patients evaluated only with confrontation visual fields were excluded; 558 additional patients were excluded because they did not have at least one follow-up visual field testing performed in our unit. The patients were evenly divided between men and women; most of the patients were white (table 1). The mean age was 48.6 years and ranged from 5 to 89 years. The median time from injury to initial formal visual field testing was 2 months. Nine patients had bilateral HH and results are presented for 263 HH. The types of visual field defects, causes of HH, and the lesion location determined by neuroimaging are detailed in table E-1 and are similar to previous reports.
Table 1 Patient characteristics (n = 254)
Changes in visual field defects.
Most (73%) of the cases had a single follow-up examination (table 2). Of the 103 cases with an examination at 3 months after the initial test, 41 (40%) had improvement in the visual field since compared to the initial test (see table 2). At 6 months after the initial test, 44 of 118 cases (37%) had improvement (see table 2). Results at other specific time points after the initial test are not shown because of insufficient numbers of cases. The median time from the initial test to the last available follow-up examination was 6 months (see table 2). At the last available follow-up examination, improvement of the visual field compared to the initial test was found in 101 HH (38.4%), including 14 HH (5.3%) that completely resolved; nonimprovement was found in 162 (61.6%) (see table 2).
Table 2 Follow-up examinations and change in visual field defect
As shown in the figure and table 3, the likelihood of spontaneous recovery decreased with increasing time from injury to the patient’s initial visual field testing. Among 39 cases first tested within 2 weeks of injury, 24 (62%) experienced improvement during subsequent follow-up. The percent of cases improving decreased to 51% for cases initially tested 2 weeks to 1 month after injury, 40% for cases tested 1 to 2 months after surgery, and 20% of cases initially tested at longer time intervals after injury. This trend for decreasing likelihood of improvement with increasing time since injury was significant (p < 0.0001, see table 3). In another analysis, cases were grouped into three intervals for the time from injury to initial testing (≤1 month, >1–2 months, and >2 months). The percent of cases showing improvement during subsequent follow-up for each of the groups was as follows: ≤1 month: 55% (62/113); >1–2 months: 40% (17/42); >2 months: 21% (22/106). The percent of cases with improvement did not differ between the ≤1 month and >1– 2 months groups (p > 0.05); however, both of these groups were different from the >2 months group with respect to the percent of cases with improvement (p < 0.05).
Figure. Graph showing the probability of improvement vs time since injury. Only cases initially tested within 6 months after the injury are included (83% of the 263 patients seen in follow-up). The estimated logistic regression function is 1/(1+exp(0.3657 – 0.36449 × Time).
Table 3 Change in visual field test according to time from injury to initial test
The figure shows the probability of improvement as a function of time since injury estimated from a logistic regression model that included cases with initial testing within 6 months of injury. The probability of improvement was related to the time since injury (p = 0.0003) with a 50 to 60% chance of improvement for cases tested within 1 month after injury that decreased to about 20% for cases tested at 6 months after surgery.
Among the 45 HH seen more than 6 months after injury, 9 HH subsequently improved (see table 3). In addition, some patients who had follow-up visual fields later than 6 months after the onset of injury showed persistent improvement after 6 months (see table 2). However, in all these cases, the visual field improvement was subtle and was associated either with improvement of the underlying disease process (such as demyelinating disease or tumor) or with improvement in the patient’s cognitive and neurologic function with better ability to perform visual field testing reliably. None of the cases with stable underlying brain disease and stable neurologic status continued to improve after 6 months.
There was no factor, other than the time from injury to initial testing, that was significantly associated with improvement. We compared the characteristics of the HH that improved over time with those that did not and did not find any significant differences between the two groups (data not shown). Similar results were found when the same comparisons were performed on the 113 HH evaluated within 4 weeks of brain injury (table 4) and on the 155 HH evaluated within 2 months of brain injury.
Table 4 Comparison of characteristics according to improvement of visual field defect for cases initially tested within 1 month of injury
Discussion.
Almost 40% of the 263 HH included in our study spontaneously improved. The characteristics of these patients and the types of visual field defects are similar to previous reports (table E-1).5–14 Unlike most previous reports, which evaluated HH recovery in primarily stroke patients, our patients had a variety of brain lesions causing their visual field defect, and are representative of the population of patients with HH (see in 16).
Although this is the largest series of HH with detailed follow-up evaluation, our results must be interpreted with caution. Indeed, our relatively low percentage of spontaneous improvement (40%) is undoubtedly biased by the fact that a number of these patients were probably seen in our service after they had already spontaneously improved. Indeed, when one considers only the group of 113 HH diagnosed within 1 month of injury, the percentage of improvement was 55% (see table 4). The figure and table 3 illustrate this finding and show that the probability of spontaneous improvement of HH decreases with the time to the patient’s first evaluation. Interestingly, there was a highly significant difference between the outcome of HH diagnosed within 2 months after injury and those diagnosed later than 2 months after injury. This supports the findings of the only two prospective studies evaluating patients with HH related to a stroke,12,13 which suggested that HH mostly recovers within the first weeks after cerebral injury. In another study,12 67% of HH had recovered within 4 weeks after a cerebral infarction. The apparent cutoff of 2 months after injury suggested by our study is consistent with these results, but is much earlier than the classic 6 months to 1 year often told to patients.11,15 Another explanation for this finding is that the patients who did not have formal visual fields within the first few weeks of their injury may have had a more devastating injury with longer hospital and intensive care unit admission, and, therefore, worse overall prognosis. However, the number of patients with associated neurologic deficits was not higher in the group of HH without improvement compared with those with improvement.
Comparison of the rate of visual field improvement at each time period after injury showed that 50 to 60% of HH seen within 1 month of injury had already improved at 3 months, while the percentage of HH with improvement between 3 and 6 months was only about 20%. Interestingly, no patient, lesion, or visual field characteristic was found to correlate with the final outcome in the 113 HH seen within 4 weeks of injury (see table 4).
Our study suggests that spontaneous improvement of HH is seen in at least 50% of patients seen within 1 month of injury. In most cases of HH, improvement occurs within the first 3 months after injury, and it is likely that most spontaneous improvement occurs very early, within the first few weeks. Spontaneous improvement later than 6 months after injury should be interpreted with caution as it is most likely related to improvement of the underlying disease process or to improvement of the patients’ ability to perform visual field tests reliably. Except for rare cases with completely reversible underlying diseases such as multiple sclerosis, the only factor associated with a higher chance of spontaneous improvement of HH is the time between the injury and evaluation. No other factor allows prediction of chances of spontaneous recovery. These results should be helpful when evaluating the efficacy of visual rehabilitation techniques. Visual field rehabilitation strategies should most likely be initiated early after injury. Furthermore, given that true spontaneous improvement after 6 months would be unusual in static neurologic diseases, documented improvement in the visual fields of these patients undergoing rehabilitative therapy would be considered a sign of therapeutic efficacy.
Footnotes
-
Additional material related to this article can be found on the Neurology Web site. Go to www.neurology.org and scroll down the Table of Contents for the March 28 issue to find the title link for this article.
See also page 906
Supported in part by a departmental grant (Department of Ophthalmology) from Research to Prevent Blindness, Inc., New York, NY, and by core grant P30-EY06360 (Department of Ophthalmology) from the National Institutes of Health, Bethesda, MD. N.J.N. is a recipient of a Research to Prevent Blindness Lew R. Wasserman Merit Award. X.Z. was supported by unrestricted educational grants from NovaVision and TEVA Neuroscience.
Disclosure: Dr. Newman is on the scientific advisory board of NovaVision.
Received July 20, 2005. Accepted in final form November 30, 2005.
References
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Hastening the Diagnosis of Amyotrophic Lateral Sclerosis
Dr. Brian Callaghan and Dr. Kellen Quigg