Predicting the Magnitude of Functional and Structural Damage in Glaucoma From Monocular Pupillary Light Responses Using Automated Pupillography.

PURPOSE

To predict the magnitude of functional damage [mean deviation (MD) on visual field examination] and structural damage [retinal nerve fiber layer (RNFL) and ganglion cell complex (GCC) thickness on spectral domain optical coherence tomography] in glaucoma from monocular pupillary light response measurements using automated pupillography.

METHODS

In total, 59 subjects (118 eyes) with either a confirmed or suspected diagnosis of glaucoma underwent automated pupillography, along with visual fields and spectral domain optical coherence tomography examinations. Association between pupillary light response measurements of each eye [amplitude of constriction, latency of onset of constriction (Loc), latency of maximal constriction (Lmaxc), velocity of constriction and velocity of redilation] and corresponding MD, average RNFL, and average GCC measurements were evaluated using univariate and multivariate regression analysis after accounting for the multicollinearity. Goodness of fit of the multivariate models was evaluated using coefficient of determination (R).

RESULTS

Multivariate regression models that contained Loc and Lmaxc showed the best association with MD (R of 0.30), average RNFL thickness (R=0.18) and average GCC thickness (R=0.26). The formula that best predicts the MD could be described as: MD=-14.06-0.15×Loc+0.06×Lmaxc. The formula that best predicts the average RNFL thickness could be described as: Average RNFL thickness=67.18-0.22×Loc+0.09×Lmaxc.

CONCLUSIONS

Glaucomatous damage as estimated by MD, RNFL, and GCC thickness measurements were best predicted by the latency parameters (Loc and Lmaxc) of pupillography. Worsening of glaucomatous damage resulted in a delayed onset of pupillary constriction and a decreased Lmaxc.