Physiological evidence for impairment in autosomal dominant optic atrophy at the pre-ganglion level.

BACKGROUND: Functional studies in patients with autosomal dominant optic atrophy (ADOA) are usually confined to analysis of physiological and clinical impact at the ganglion cell (GG) and post GC levels. Here we aimed to investigate the impact of the disease at a pre-GC level and its correlation with GC/post-GC related measures. METHODS: Visual function was assessed in a population of 22 subjects (44 eyes) from 13 families with ADOA submitted to OPA1 mutation analysis. Quantitative psychophysical methods were used to assess konio and parvocellular chromatic pathways (Cambridge Colour Test) and distinct achromatic spatial frequency channels (Metropsis Contrast Sensitivity Test). Preganglionic and GC measures were assessed with the Multifocal Electroretinogram (mfERG) and Pattern Electroretinogram (PERG) respectively. Global Pattern and Multifocal VEP (visual evoked potentials) were used to assess retinocortical processing, in order to characterize impaired processing at the post GC level. Perimetric sensitivity, retinal and ganglion cell nerve fibre layer (RNFL) thickness measurements were also obtained. RESULTS: Chromatic thresholds were significantly increased for protan, deutan and tritan axes (p < < 0.001 for all comparisons) and achromatic contrast sensitivity (CS) was reduced for all studied six spatial frequency channels (p < < 0.001). We observed significant decreases in peripapillary (p ≤ 0.0008), macular (ring2: p = 0.02; ring 3: p < 0.0001) RNFL, as well as in overall retinal thickness (p < 0.0001 in all regions, except the central one). Interestingly, we found significant decreases in pre-ganglionic multifocal ERG response amplitudes (P1-wave: p ≤ 0.005) that were correlated with retinal thickness (ring 2: r = 0.512; p = 0.026/ring 3: r = 0.583; p = 0.011) and visual acuity (r = 0.458; p = 0.03, central ring 1). Reductions in GC and optic nerve responses amplitude (PERG: p < 0.0001, P50 and N95 components; Pattern VEP: p < 0.0001, P100) were accompanied by abnormalities of the MfVEP, primarily in central locations (ring 1: p = 0.0007; ring 2: p = 0.012). CONCLUSIONS: In the ADOA model of ganglion cell damage, parvo-, konio- and magnocellular pathways are concomitantly affected. Structural changes and physiological impairment also occurs at a preganglionic level, suggesting a retrograde damage mechanism with a significant clinical impact on visual function, as shown by correlation analysis. Cortical impairment is only moderately explained by the retinal phenotype, suggesting additional damage mechanisms at the cortical level.