Objectives: To objectively assess pupillary involvement according to various etiologies of acquired isolated third nerve palsy using automated pupillometry, and evaluate the efficacy of digital pupillometry in discriminating compressive lesions from microvascular ischemic third nerve palsy.
Design: Retrospective, observational case series.
Methods: A total of 171 subjects were included in this study, consisting of 60 subjects with presumed microvascular ischemic third nerve palsy, 51 with non-ischemic third nerve palsy, and 60 controls whose pupillary light responses were measured using a dynamic automated pupillometer. Subjects with non-ischemic third nerve palsy were divided into subgroups according to their etiology; inflammatory and compressive groups including tumor and aneurysm. Pupillometry parameters including minimum and maximum pupil diameters, constriction latency and ratio, maximum and average constriction velocities and dilation velocity were noted. The diagnostic ability of pupillometry parameters for discriminating compressive vs microvascular ischemic third nerve palsy was evaluated. The inter-eye difference of the involved eye and the uninvolved fellow eye was calculated to adjust for individual variability.
Results: Among all parameters, reduced pupillary constriction ratio was the most specific parameter for detecting non-ischemic third nerve palsy, as a large inter-eye difference beyond the normative range of controls was found in 0% of ischemic, 20% of inflammatory and 60% of compressive third nerve palsy. With the diagnostic criteria using inter-eye differences of 1) minimum pupil diameter > 0.45 mm, or 2) pupillary constriction ratio < -7.5% compared to the fellow eye, the sensitivity and specificity for diagnosing compressive third nerve palsy were 95% and 88%, respectively. In the compressive group, positive correlations were found between the degree of external ophthalmoplegia and constriction ratio (r = 0.615, p<0.001), average constriction velocity (r = 0.591, p = 0.001) and maximum constriction velocity (r = 0.582, p = 0.001).
Conclusions: Abnormal pupillary constriction ratio was highly specific for detecting compressive third nerve palsy, although the sensitivity was not high. Digital pupillometry demonstrated relatively good performance for discriminating compressive lesions from microvascular ischemic third nerve palsy.