Novel Structure-Function Models for Estimating Retinal Ganglion Cell Count Using Pattern Electroretinography in...

Background/Objectives: The early detection of retinal ganglion cell (RGC) dysfunction is critical for timely intervention in glaucoma suspects (GSs). The combined structure-function index (CSFI), which uses visual field and optical coherence tomography (OCT) data to estimate RGC counts, may be of limited utility in GSs. This study evaluates whether steady-state pattern electroretinogram (ssPERG)-derived estimates better predict early structural changes in GSs. Methods: Fifty eyes from 25 glaucoma suspects underwent ssPERG and spectral-domain OCT. Estimated RGC counts (eRGCC) were calculated using three parameters: ssPERG-Magnitude (eRGCC_Mag), ssPERG-MagnitudeD (eRGCC_MagD), and CSFI (eRGCC_CSFI). Linear regression and multivariable models were used to assess each model's ability to predict the average retinal nerve fiber layer thickness (AvRNFLT), average ganglion cell layer-inner plexiform layer thickness (AvGCL-IPLT), and rim area. Results: eRGCC_Mag and eRGCC_MagD were significantly correlated with eRGCC_CSFI. Both PERG-derived models outperformed eRGCC_CSFI in predicting AvRNFLT and AvGCL-IPLT, with eRGCC_MagD showing the strongest association with AvGCL-IPLT. Conversely, the rim area was best predicted by eRGCC_Mag and eRGCC_CSFI. These findings support a linear relationship between ssPERG parameters and early RGC structural changes, while the logarithmic nature of visual field loss may limit eRGCC_CSFI's predictive accuracy in GSs. Conclusions: ssPERG-derived estimates, particularly eRGCC_MagD, better predict early structural changes in GSs than eRGCC_CSFI. eRGCC_MagD's superior performance in predicting GCL-IPLT highlights its potential utility as an early biomarker of glaucomatous damage. ssPERG-based models offer a simpler and more sensitive tool for early glaucoma risk stratification, and may provide a clinical benchmark for tracking recoverable RGC dysfunction and treatment response.