Abstract
Purpose:
To analyze the utility of automated macular ganglion cell layer (GCL) measurements in patients with radiologic evidence of chiasmal compression. To compare GCL measurements to retinal nerve fiber layer (RNFL) thickness, macular thickness and visual fields in affected patients.
Methods:
Patients with active chiasmal compression on MRI seen between January 2010 and October 2013 were eligible. Age-matched controls were also included. All were studied with the Cirrus HD-OCT macular cube 512x128 and RNFL scan protocols, as well as 30-2 visual fields. One eye was selected at random per person for analysis. T-tests were done comparing RNFL, GCL and macular thicknesses between patients and controls. Effect size (d) was calculated to assess the magnitude of difference between them. Average GCL, RNFL and macular thicknesses were correlated to mean deviation (MD).
Results:
Twenty-six eyes from 13 patients and 13 controls were enrolled. Ten patients had pituitary adenoma, 2 had tuberculum sella meningioma and 1 had Rathke’s cyst. Six patients had bitemporal hemianopia, 5 did not have field defects and 2 had junctional scotomas. Average GCL thickness was 70±8 μm in patients and 80±5 μm in controls (p< 0.001). Average macular thickness was 268±13 μm in patients and 278±9 μm in controls (p= 0.03). Average RNFL thickness was 84±13 μm in patients and 90±6 μm in controls (p= not significant). Effect size was greatest for GCL thickness (d= 1.44), followed by macular thickness (d= 0.91) then RNFL thickness (d= 0.64). MD was more correlated to GCL thickness (r2= 0.14) than macular (r2= 0.005) or RNFL thicknesses (r2= 0.05). One patient with follow-up showed baseline nasal GCL thinning, bitemporal field defects, but normal RNFL and macular thicknesses. Post-resection, visual fields normalized with stable nasal GCL thinning.
Conclusions:
The greatest difference in thickness measurements between patients and controls was in the GCL, followed by macula and least in RNFL, reflected by both p-values and effect sizes. Thinning of the GCL may be detected prior to significant loss of RNFL or macular thickness in these patients. Since GCL thinning is present in patients with normal visual fields after chiasmal decompression, and since it is a more objective measurement, GCL analysis may be the best metric to follow patients with chiasmal lesions.
Keywords: 550 imaging/image analysis: clinical •
629 optic nerve •
531 ganglion cells