controlRedStimulus.

Purpose
To test whether differences in pupil responses to melanopsin-isolating spectral stimuli in glaucoma may be useful as a diagnostic biomarker.
Methods
Spectral stimuli were presented to 20 glaucoma and 15 age-similar healthy control participants. Stimuli were pairs of silent-substitution spectra designed to provide (1) equal stimulation to cone photoreceptors but maximum (~325%) contrast to melanopsin or (2) equal stimulation to melanopsin but ~325% contrast to cones. Narrowband long-wavelength/red (657 nm) and short-wavelength/blue (471 nm) pulses were also presented from a dark background to 16 glaucoma and 12 control participants. Pulses lasted 3 seconds and pupil size was measured for 15 seconds. Pupil response metrics were compared by t-test and relationships with visual field and OCT summary indices were assessed by Spearman’s rank correlation. Diagnostic accuracy was measured by area under the receiver operating characteristic curve (AUC).
Results
Pupil constriction was more persistent after pulse offset for the melanopsin-directed stimulus (2% mean paired difference 6s post-pulse offset, p < 0.001). All pupil parameters were similar between groups (p = 0.04–0.90) for all stimuli. Correlations between pupil response parameters and visual field summary indices and circumpapillary retinal nerve fibre layer thickness were weak (rho 0.02–0.57, all p > 0.05). Diagnostic accuracy for all pupil parameters was poor, with AUC 95% confidence intervals overlapping 0.5 for all but time to maximal constriction for the cone-directed stimulus.
Conclusions
Pupil responses to melanopsin-isolating spectra were similar between glaucoma and control participants. Pupillary responses to melanopsin-isolating silent substitution spectra are unlikely to be useful as a diagnostic biomarker for glaucoma.