Predicting the intereye asymmetry in functional and structural damage in glaucoma using automated pupillography

To predict the intereye asymmetry in functional (mean deviation, MD on visual field, VF) and structural (retinal nerve fibre layer, RNFL and ganglion cell complex, GCC thickness on spectral domain optical coherence tomography, SDOCT) measurements in glaucoma using the automated pupillography parameters.

Fifty-nine subjects with a diagnosis of either glaucoma or glaucoma suspect underwent automated pupillography along with VF and SDOCT examinations. Association between pupillography and the absolute intereye difference in MD, RNFL and GCC measurements was evaluated using regression analysis after accounting for the multicollinearity.

Univariate regression analysis showed statistically significant associations (p < 0.05) between multiple pupillography parameters and the intereye difference in MD, RNFL and GCC thickness measurements. Multivariate regression with less strongly correlated parameters identified intereye difference in amplitude change (Ac) per cent to be the parameter that best predicted the intereye asymmetry in MD (Intereye asymmetry in MD = 2.20 + 1.33*intereye difference in Ac per cent, R²  = 0.36), RNFL thickness (3.38 + 3.55*intereye difference in Ac per cent, R²  = 0.49) and GCC thickness (4.49 + 2.06* intereye difference in Ac per cent, R²  = 0.41). Ability of intereye Ac per cent difference to predict intereye asymmetry in MD, RNFL and GCC thickness was better in patients with angle closure disease (R²  = 0.38, 0.79, 0.66, respectively) compared to those with open angles (R²  = 0.25, 0.15, 0.16, respectively).

Intereye asymmetry in MD, RNFL and GCC thickness measurements was best predicted by the intereye difference in Ac per cent on automated pupillography. The predicting ability was better in patients with angle closure compared to those with open angles.