Correlation between oscillatory potentials and photopic negative response amplitudes in a clinical setting

Abstract

Purpose : According to data presented before (IOVS 2019 v.60:5969), a strong correlation between the timing of retinal ganglion cell bursts and oscillatory potential (OP) peaks was observed when bright white flashes were presented in rats. Based on that, a hypothesis can be put forward about the existence of a correlation between OPs and the photopic negative response (PhNR) of the full-field electroretinogram (ERG). The purpose of the current study was to compare OP amplitude and PhNR parameters in patients with various retinal pathologies.

Methods : A retrospective chart review and data analysis of patients aged 18 and over undergoing routine ERG testing at USF Eye Institute (Tampa, FL) was conducted. The Photopic 3.0 ERG responses (LA 3.0) were analyzed. Ph3 responses were recorded using DTL electrodes, white xenon flash (2.5 cd.s/m²) on a white background (30 c/m²) with a digitization rate of 3750 Hz and bandpass filtering range of 0.3 to 1500 Hz. The ERG signal was filtered using high-pass 4^th order Butterworth filter with 58 Hz cut-off frequency. The root means square (RMS) of the filtered signal was calculated for the OP portion of the signal. The OP RMS values were compared (linear regression) to the amplitude of the PhNR response, determined in 3 different ways: measured from the isoelectric baseline (fB), measured from the b-wave peak (fP) and as an area under the curve (AUC).

Results : The records of 11 patients/20 eyes (2M, 9F); mean age 46.7 +/-14.6 yrs. were evaluated. The correlation between OP RMS and PhNR(fB) was moderate at R²=0.5515 (p<0.001), while the correlation with PhNR(fP) was considerably better at R²=0.8222 (p<0.001). Furthermore, several models of a linear regression between OP RMS and AUC extended to different time points after b-wave peak (every 5 ms form 5 ms to 50 ms after b-wave peak), which indicated a good correlation across the range (R² 0.82 to 0.89) with the best regression model at 15 ms after b-wave peak (R² = 0.8875, p<0.0001).

Conclusions : A reasonably strong correlation was observed between OP parameters and PhNR amplitude. That supports a relationship between OP amplitude and PhNR under LA 3.0 recoding conditions, which has applicability in clinical practice, especially when determining the amplitude of PhNR with traditional methods is unreliable.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.