Efficacy of Visual Retraining in the Hemianopic Field after Stroke: Results of a Randomized Clinical Trial

• Published in: Ophthalmology (Rochester, Minn.) Vol. 128; no. 7; pp. 1091 - 1101
• Main Authors: Cavanaugh, Matthew R, Blanchard, Lisa M, McDermott, Michael, Lam, Byron L, Tamhankar, Madhura, Feldon, Steven E
• Format: Journal Article
• Language: English
• Published: United States 01-07-2021
• Subjects: Clinical trials; Stroke; Visual loss; Neuro-ophthalmology; Visual fields
• ISSN: 1549-4713
• DOI: 10.1016/j.ophtha.2020.11.020

To evaluate the efficacy of motion discrimination training as a potential therapy for stroke-induced hemianopic visual field defects.
Clinical trial.
Forty-eight patients with stroke-induced homonymous hemianopia (HH) were randomized into 2 training arms: intervention and control. Patients were between 21 and 75 years of age and showed no ocular issues at presentation.
Patients were trained on a motion discrimination task previously evidenced to reduce visual field deficits, but not in a randomized clinical trial. Patients were randomized with equal allocation to receive training in either their sighted or deficit visual fields. Training was performed at home for 6 months, consisting of repeated visual discriminations at a single location for 20 to 30 minutes daily. Study staff and patients were masked to training type. Testing before and after training was identical, consisting of Humphrey visual fields (Carl Zeiss Meditech), macular integrity assessment perimetry, OCT, motion discrimination performance, and visual quality-of-life questionnaires.
Primary outcome measures were changes in perimetric mean deviation (PMD) on Humphrey Visual Field Analyzer in both eyes.
Mean PMDs improved over 6 months in deficit-trained patients (mean change in the right eye, 0.58 dB; 95% confidence interval, 0.07-1.08 dB; mean change in the left eye 0.84 dB; 95% confidence interval, 0.22-1.47 dB). No improvement was observed in sighted-trained patients (mean change in the right eye, 0.12 dB; 95% confidence interval, -0.38 to 0.62 dB; mean change in the left eye, 0.10 dB; 95% confidence interval, -0.52 to 0.72 dB). However, no significant differences were found between the alternative training methods (right eye, P = 0.19; left eye, P = 0.10).
To date, no widely accepted therapy is available to treat HH. This study evaluated the efficacy of a promising potential treatment, visual perceptual training. We failed to find a difference between treatment training within the deficit field and control training within the sighted field when performed in a home environment.