Vision restoration therapy (VRT) efficacy as assessed by comparative perimetric analysis and subjective questionnaires.

PURPOSE: We wished to evaluate the efficacy of vision restoration therapy (VRT) in patients with post-chiasmatic brain damage using different functional perimetric tests. These were compared with measures of subjective vision and reaction time.
METHODS: An open trial was conducted with hemianopia/scotoma (n=16) patients. Before and after 6 months of VRT results of high resolution (HRP) and Tuebingen automated perimetry (TAP) were evaluated and compared to performance in a Scanning Laser Ophthalmoscope (SLO) as previously reported. Whereas TAP and HRP used above-threshold or near-threshold individual target stimuli on grey background, the SLO used a psychophysical task of detection of three black targets (reverse stimulus) on bright red, patterned background. Subjective testimonials of activities of daily living (ADL) were probed with questionnaires and interviews.
RESULTS: Before VRT, the visual field border as assessed by SLO was located significantly closer to the vertical midline than the HRP and TAP border (border mismatch). After VRT the SLO border was still unchanged whereas HRP measurements revealed significant border shifts due to improved stimulus detection (p<0.0001) and improved reaction time (p<0.005) . Fewer misses were also observed in both eyes with TAP (p<0.01) which was primarily due to a significant shift of the absolute borders. Thus, VRT potentiated the mismatch between the SLO borders and the HRP/TAP borders. Fixation performance and the blind spot position remained unchanged after VRT. ADL ratings in the questionnaire improved significantly after VRT which was confirmed by independent patient testimonials.
CONCLUSIONS: We replicated earlier findings that VRT improves stimulus detection in HRP and TAP perimetry which were accompanied by subjective, visual improvements. These changes are not caused by fixation or eye movement artifacts. Because the SLO border was located significantly closer to the vertical midline before VRT ("border mismatch") and, in contrast to HRP and TAP, did not change after VRT, we interpret this border mismatch to indicate that the SLO task was too difficult to perform and thus insensitive to VRT effects. Significant reaction time improvements indicate that plasticity of temporal processing might play an important role in vision restoration after brain damage. A further description of the precise psychophysical nature of the restored areas of residual vision is now warranted.