Onset and Progression of Retinal Nerve Fiber Layer (RNFL) Damage in Experimental Glaucoma Occur Earlier by Scanning Laser Polarimetry (SLP) than by Spectral Domain Optical Coherence Tomography (SDOCT)

摘要

Purpose: : To test the hypothesis that axonal cytoskeletal disruption precedes glaucomatous axon loss, longitudinal SLP and SDOCT measurements of the peripapillary RFNL were compared in terms of time to reach onset of damage and time to reach a specific progression endpoint. Methods: : This study included 33 rhesus macaques with unilateral experimental glaucoma (EG). Each had three or more weekly baseline measurements in both eyes of peripapillary RNFL thickness by SDOCT (Spectralis, Heidelberg Engineering, GmbH) and RNFL retardance by SLP (GDxVCC, Carl Zeiss Meditech, Inc). Laser photocoagulation was then applied to the trabecular meshwork of one eye to induce chronic elevation of IOP and weekly imaging continued. Pairwise differences between baseline observations were sampled by bootstrapping to determine the 95% confidence limits of each measurementa??s repeatability. The first two sequential measurements below the lower confidence limit defined the endpoint for each test. McNemara??s test and survival analysis were applied to compare endpoint proportions and times to endpoint, respectively. For all EG eyes reaching endpoint by both SLP and SDOCT, bilinear (segmental) regression and exponential decay functions were fit to the RNFL-vs-time series to determine the time to damage onset; paired t-tests were used to compare fit-derived parameter differences. Results: : A larger proportion of EG eyes reached endpoint first by SLP (N=20, 61%) than did by SDOCT (N=6, 18%; ?§2=6.5, two-tailed p=0.01, McNemara??s test); seven did not reach endpoint by either RNFL measure before being sacrificed for other studies. Survival analysis indicated a shorter time to endpoint for SLP than for SDOCT (?§2=7.7, p=0.005, Log-rank Mantel-Cox Test and ?§2=6.4, p=0.01, Gehan-Breslow-Wilcoxon Test). Of the 16 EG eyes that reached endpoint by both measures, the median times to endpoint were 99.5 and 172 days for SLP and SDOCT, respectively (p=0.0087). The time to onset was faster for SLP than for SDOCT based on either bilinear segmental fits (by 28 days; p=0.006, average R2 =0.81) or exponential fits (by 36 days; p=0.01, average R2 =0.89). There were no goodness-of-fit differences between SLP and SDOCT for either model (p=0.44 and p=0.34, respectively). Specificity assessed in controls was 91% for SLP and 100% for SDOCT. Conclusions: : This study demonstrates that the onset of glaucomatous RNFL damage measured by SLP occurs earlier than RNFL thinning measured by SDOCT. These results are consistent with our previous reports and the hypothesis that within the RNFL axonal cytoskeletal disruption precedes glaucomatous axon loss.