Examination of retinal structure and visual function in civilian patients with Traumatic Brain Injury

Abstract

Purpose: Recent reports suggest that veterans with traumatic brain injury (TBI) may have optic nerve head (ONH) and inner retinal damage in addition to reduced visual performance. We seek to (1) quantify the extent to which abnormalities exist in retinal/ONH structure and visual function in civilian TBI patients and (2) correlate structural and functional measurements in the same eyes.

Methods: Spectral domain optical coherence tomography volume scans of the ONH and macula were acquired in 14 eyes of 7 TBI patients (36 ± 12 years). Peripapillary retinal nerve fiber layer [RNFL] and macular retinal ganglion cell-inner plexiform layer [GCIPL] thicknesses were quantified globally and in sectors, and compared with instrument-based normative data. Functional measurements were acquired with the full-field flash electroretinogram (to quantify photopic negative response [PhNR] amplitude) and 30-2 standard automated perimetry (to quantify mean deviation [MD] and mean sensitivity). The percentage of agreement between global structural and functional measures (i.e., whether compared measures were both normal or abnormal) was calculated, as was the probability of agreement when corrected for chance (AC1 statistics). Structural and functional measures were correlated on global and local scales via regression analyses.

Results: GCIPL and RNFL thickness abnormalities were detected in 8/14 and 5/14 eyes, respectively. Abnormalities in PhNR amplitude compared to normative values were found in 12/14 eyes and in MD in 10/14 eyes. The strongest agreements occurred between PhNR and MD (64% of eyes, AC1=0.42), GCIPL and RNFL thickness (64%, AC1=0.31), and RNFL thickness and MD (64%, AC1=0.29). Significant linear relationships (P<.05) were found across eyes between GCIPL and RNFL thicknesses (R² =0.62), MD and GCIPL thickness (R² =0.49), and MD and RNFL thickness (R² =0.63). Only 9/56 comparisons were statistically significant when using published structure-function correspondence maps to correlate sector measures of GCIPL/RNFL thickness with corresponding local measures of visual field sensitivity.

Conclusions: While structural abnormalities exist in some eyes, our data indicate a high prevalence of functional abnormalities in TBI patients. A better understanding of retinal/ONH structure, visual function and their correlation in TBI patients may enable more effective diagnosis, classification and treatment of injury.