Luminance and colour variant pupil perimetry in glaucoma.

BACKGROUND: This study investigated the diagnostic utility for glaucoma of multifocal pupillographic objective perimetry stimuli targeting different components of the pupillary response: cortically derived colour responses and subcortical luminance responses.
DESIGN: Observational cross-sectional study undertaken at the Australian National University. PARTICIPANTS: Thirty-five eyes of 24 glaucoma subjects and 46 eyes of 23 normal subjects.
METHODS: Subjects were tested with three multifocal pupillographic objective perimetry stimulus variants. The luminance-only variant (YYbal) utilized yellow stimuli on a yellow background; mixed colour and luminance protocols utilized green stimuli on a red background (RGbal, RG). Stimuli of 33 ms duration were presented at mean intervals of 4 s/region. MAIN OUTCOME MEASURES: Pupil constriction amplitude and time to peak. Area under the receiver operating characteristic curve was the main measure of sensitivity and specificity for glaucoma.
RESULTS: Colour and luminance protocols were more accurate at differentiating glaucoma subjects from normal subjects than the luminance-only protocol, and produced the largest reductions in amplitudes. This type of protocol also produced the highest overall sensitivity and specificity for glaucoma (receiver operating characteristic % area under the curve: severe, 100%; moderate, 94.4%; mild, 71.0%). Pattern deviations tended to produce higher area under the receiver operating characteristic curves in eyes classified as mild. Significant differences in the means of the six worst amplitude deviations were observed between normal and severe glaucoma subjects only.
CONCLUSIONS: Stimuli targeting both cortical pupillary colour response and subcortical pupillary luminance response components produced higher diagnostic accuracy than stimuli targeting subcortical pupillary luminance responses alone. Inclusion of constriction latencies further improved accuracy.