Retinal Ganglion Cell Topography in Patients With Visual Pathway Pathology

To investigate and quantify the impact of intracranial lesions at different locations within the visual pathway on the ganglion cell layer-inner plexiform layer (GCL-IPL) complex and the retinal nerve fiber layer (RNFL). Patients with intracranial lesions affecting the optic chiasm (Group I) or the optic tract and/or lateral geniculate nucleus (Group II) were included. All patients received kinetic visual field assessment and underwent spectral domain optical coherence tomography. Peripapillary and papillomacular bundle (PMB) RNFL and macular GCL-IPL thickness in 4 perifoveal areas were measured and compared with normal values derived from 52 age-matched healthy control subjects. Z-scores for each parameter of every patient were calculated and compared with the normative data. Z-scores less than -2.0 (e.g., -2.5) were considered as being statistically significant. Twenty-two patients (Group I and II: 13 and 9, respectively) were included. Ten of 13 patients in Group I showed significant binasal GCL-IPL thinning, with associated temporal sector thinning in 8 patients. In Group II, all 9 patients showed significant reduction of the GCL-IPL corresponding to the homonymous visual field defect, but only 4 demonstrated RNFL thinning. Contralateral RNFL thinning within the PMB clinically similar to bow-tie atrophy was evident in all patients in Group II. GCL-IPL and RNFL thinning varied in severity from mild (isolated PMB RNFL thickness reduction) to severe (bilateral asymmetrical reduction of PMB RNFL associated with asymmetric, predominantly nasal reduction of GCL-IPL) in Group I. Clinical abnormalities in patients with visual pathway lesions are more likely to demonstrate abnormalities of GCL-IPL than global peripapillary RNFL thickness. However, PMB thickness measurement appears to be a valuable tool to detect abnormalities of the anterior visual pathways. If peripapillary RNFL measurements are performed in such patients, PMB thickness should be considered the most useful quantitative parameter.