Abstract
Purpose:
To relate RGC functional loss and RNFL retardance changes to progressive loss of RNFLT.
Methods:
N=36 rhesus macaques with unilateral experimental glaucoma (EG) each had ≥3 baseline (BL) measurements of peripapillary RNFLT (Spectralis SDOCT, Heidelberg Engineering, GmbH), retardance (GDxVCC, Carl Zeiss Meditec, Inc) and multifocal electroretinography (mfERG, VERIS, Electro-Diagnostic Imaging, Inc). mfERG responses were high-pass filtered (>75 Hz) to measure high- and low-frequency component (HFC and LFC) amplitudes, including LFC peaks N1, P1 and N2. Chronic unilateral intraocular pressure (IOP) elevation was induced by trabecular meshwork laser; alternating bi-weekly testing continued (SDOCT one week, SLP and mfERG the next). The measurement noise (95% confidence interval, CI) of each parameter was determined by bootstrapping all BL observations. Onset of RNFLT change was defined as the first of two sequential observations below the noise limit.
Results:
Post-laser follow-up was 9±7 months; mean and peak IOP were 18±5 and 39±12 mmHg in EG eyes, 11±2 and 17±6 mmHg in fellow control eyes (CTL). At the final available time point, RNFLT had decreased from BL by 13±13%, retardance by 19±11% and the mfERG HFC by 29±17% (p<0.0001 each); CTL eyes changed by +0.3±3% (p=0.44), +0.7±6%, (p=0.45) and -6±14% (p=0.01), respectively. Retardance and HFC were correlated with RNFLT at the final time point (R2=0.57, p<0.0001 and R2=0.37, p<0.0001, respectively) but none of the LFC amplitudes were correlated with RNFLT (N1: R2=0.08, p=0.1; P1: R2=0.04, p=0.3; N2: R2=0.08, p=0.1). A linear model was most appropriate (F-test, p≥0.6 each) to relate relative loss (final:BL) of each parameter to RNFLT loss; the resultant slope was >0 for retardance (0.65, p<0.0001) and HFC (0.72, p=0.001) but not for any LFC parameter; the Y-intercept was significantly negative for retardance (-11%, 95% CI: -7% to -15%) and mfERG HFC (-20%, 95% CI: -13% to -27%). Onset of RNFLT change occurred in 25/36 EG eyes and 0/36 CTL eyes; at the prior test session (median 8 days earlier), retardance was reduced by 17±10% and HFC by 28±13% below BL (p<0.0001 each).
Conclusions:
RGC axonal cytoskeletal disruption and specific RGC functional changes, as measured by RNFL retardance and mfERG, respectively, are present at the onset of RNFLT change in EG and correlate with its progressive loss beyond onset.
Keywords: 610 nerve fiber layer •
507 electrophysiology: clinical •
552 imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)