Chen and associates reported on the efficacy and safety of apraclonidine 0.5% in diagnosing Horner syndrome in children under high and low illumination. They describe a prospective study comparing 10 patients with Horner syndrome (confirmed with positive cocaine 4% test results) with 10 patients with physiologic anisocoria (with negative cocaine 4% test results). The authors conclude that in the Horner syndrome group, apraclonidine 0.5% gives a reversal of anisocoria in all patients in light conditions, but in only 7 patients in dim conditions. Therefore, they posit that it is effective as a diagnostic method in diagnosing Horner syndrome, especially in light conditions. They also conclude that apraclonidine is a safe drug to in to use in the diagnosis of Horner syndrome in children.

The authors analyzed the data comparing the differences in size of the smaller pupil compared with the larger pupil for different light illuminations, with or without apraclonidine 0.5%. We analyzed their data differently, looking at the differences in each pupil size in high levels of illumination compared with low levels, before the instillation of apraclonidine. The same analysis was applied to pupils after the application of apraclonidine, that is, looking at the differences in pupil size in high compared to low levels of illumination. We used the Wilcoxon signed-rank test to compare groups.

We found that, before instillation of apraclonidine, the mean difference in pupil size of the 10 patients in the Horner syndrome group between high and low illumination was 0.9 mm (standard deviation [SD], 0.5 mm) compared with 1.7 mm (SD, 0.4 mm) in the physiologic pupil group. This difference was statistically significant ( P = .024). The mean difference between the groups between high and low illumination when apraclonidine was used was 1.7 mm (SD, 0.4 mm) and 1.3 mm (SD, 0.8 mm), respectively. This difference was not statistically significant ( P = .286). This demonstrates that simply comparing the size of each pupil between high and low levels of illumination is an effective method of diagnosing Horner syndrome without using any pharmacologic agents.

There have been recent reports of serious idiosyncratic side effects of apraclonidine, especially in children younger than 6 months of age. Side effects include excessive drowsiness, bradycardia, hypertension, and decreased oxygen saturations. We believe that the small size of Chen and associates’ study, together with the absence of children younger than 1 year of age, makes it difficult to conclude that apraclonidine is safe to use to diagnose Horner syndrome.

We advise that, where there is diagnostic uncertainty, the diagnosis of Horner syndrome be made by comparing the size of each pupil using differing light conditions (if necessary, by the photographic method described by Chen and associates ) and the tried and tested safe method of cocaine 4%.