Safety of apraclonidine eye drops in diagnosis of Horner syndrome in an outpatient pediatric ophthalmology clinic

Abstract

Purpose

To describe the efficacy and systemic side effects of apraclonidine drops 0.5% in children clinically suspected of having Horner syndrome.

Methods

The medical records of patients with clinically suspected Horner syndrome who underwent testing with apraclonidine 0.5% eyedrops were reviewed retrospectively. The following data were retrieved from the record: allergic reactions, somnolence, shallow respiration, pallor, or excessive fussiness noted by the examiner or parents.

Results

A total of 46 patients presenting with anisocoria and tested with apraclonidine 0.5% were included. Of these, 15 had a positive result, with reversal of anisocoria. The mean age of patients was 3.2 years (median, 0.91; mode, 0.25 years). Twenty-four patients were ≤1 year of age; 19 were ≤6 months. No systemic side effects were noted during the examination or reported by parents in any patients.

Conclusions

The use of topical apraclonidine eyedrops has been described as an effective test for Horner syndrome. However, concerns have been raised regarding the potential systemic side effects in children, especially those under the age of 6 months. In our cohort, no systemic side effects were reported, including in those under 6 months of age.

Horner syndrome, or oculosympathoparesis, classically presents with the triad of miosis, upper lid ptosis (together with mild elevation of the lower lid, or reverse ptosis), and anhidrosis of the ipsilateral face (the latter depending on the site of sympathetic interruption).^1,2 Pathologic causes may range from migraines to life-threatening carotid dissection or malignancy.¹ Early recognition and intervention may be lifesaving.³

Confirmation of the suspected clinical diagnosis was traditionally established by pharmacological testing of the pupils using cocaine drops.⁴ This produces an increase in the pupillary diameter in the normal eye, whereas in Horner syndrome the abnormal pupil does not dilate as well as the normal one.³ Cocaine is a controlled substance that is difficult to obtain.

Apraclonidine is an adrenergic receptor agonist used for its antiocular hypertensive effects in glaucoma. In Horner syndrome, there is reversal of the anisocoria (Figure 1) due to the alpha-2 agonist effect of apraclonidine on the normal pupil (causing it to constrict) and the dominant alpha-1 effect from adrenergic receptor supersensitivity in the Horner pupil, which takes 2-5 days to develop, causing it to dilate.^3,4,6,7 Partial resolution of the ptosis has been reported in some cases as well.^4,8

FIG 1.

Apraclonidine eyedrop test in a child with suspected right Horner syndrome. A, Pretesting with apraclonidine eyedrops. The right pupil is smaller in diameter compared with the left, which was more evident in dim light; there is mild right upper lid ptosis. B, Reversal of anisocoria after apraclonidine eye drops, with the right pupil now larger than the left and reversal of right upper lid ptosis.

Morales and colleagues⁶ described the reversal of Horner syndrome anisocoria with apraclonidine 1% eye drops. Other reports described the use of apraclonidine 0.5% in diagnosing Horner syndrome^4,7 and it is believed to have the same sensitivity and specificity as the cocaine test for diagnosing oculosympathoparesis.⁸ The ready availability of apraclonidine eye drops resulted in it being recommended as the preferred test for Horner syndrome,⁷ with some authors recommending its adoption as a gold standard in clinical practice.⁹

Concerns have been raised regarding the use of apraclonidine in young children, especially those under 6 months of age. Adverse side effects of topical administration of this drug have been described—including decreased respiratory rate, somnolence, lethargy, and bradycardia¹⁰ in some reports. The purpose of the current study was to assess the safety of the apraclonidine 0.5% eye drop test in children with anisocoria and suspected Horner syndrome and to describe the incidence of side effects, if any.

Subjects and Methods

This study was approved by the University of Pittsburgh Institutional Review Board. The medical records of children who underwent pharmacological testing with apraclonidine 0.5% at the University of Pittsburgh Medical Center (UPMC) Children’s Hospital of Pittsburgh (CHP) for suspected Horner syndrome between February 2016 and October 2020 were reviewed retrospectively. Demographic and ophthalmic data collected included age, sex, concentration of apraclonidine eye drops used, and results of drop testing. A positive test was documented when reversal of anisocoria was observed by the physician after apraclonidine eye drop testing. The medical record of each child was reviewed for any documented side effects, such as allergic reactions, somnolence, shallow respiration, pallor, or excessive fussiness documented by the physician or reported by the parent/guardian. The time spent by the child under observation was also noted. In our practice, the apraclonidine test is typically carried out in our outpatient clinicafter explaining the test and potential side effects to parents/guardians.

Drops were instilled in both eyes, one drop at a time. Punctal occlusion for 1 minute after eyedrop instillation to decrease systemic absorption of the drug was performed when possible. The test was interpreted 30-40 minutes after eye drop instillation, and the child was observed for a minimum of 40 minutes after testing to monitor for any potential side effects, usually in the immediate waiting area.

Results

We identified 46 patients (27 females) that underwent pharmacological testing with apraclonidine eye drops because of anisocoria for exclusion of Horner syndrome (Table 1). Mean patient age was 3.2 years (mode, 3 months; median, 0.91 years; range 7 weeks to 15 years)). Of the 46 children, 24 were 12 months or younger; 19 patients were 6 months or younger. Time spent under observation was clearly documented in the physician’s note for 25 patients and ranged from 40 to 140 minutes (mean, 89.1). An additional 9 patients underwent pupillary dilation and examination after their apraclonidine test, with an approximate time of at least 75 minutes spent in clinic. One patient was sent to the emergency room after 25 minutes for admission and work up after a positive test with reversal of anisocoria; this patient did not have any systemic side effects from the topical administration of apraclonidine 0.5%. Three patients were tested as inpatients while admitted for other reasons. In the remaining 8 patients, time spent under observation was unclear. None of the 46 patients had any documented topical or systemic adverse reactions. Reversal of anisocoria was noted in 15 patients, prompting further investigation for Horner syndrome. The workup was abnormal in 4 patients, the most significant of which was a 3-month-old girl in whom magnetic resonance imaging of the brain and neck revealed a left lower neck mass (3 × 2.7 × 2.7 cm) extending into the superior mediastinum that was consistent with neuroblastoma. This patient was urgently referred to oncology for further workup and management.

Table 1.

Summarizing the results of the 46 patients undergoing pharmacological testing with apraclonidine for suspected Horner syndrome

Study parameter	Result
Age, mean/mode (range)	3.2 years/3 months (7 weeks to 15 years)
Sex, no.
Male	19
Female	27
Result of testing, no.
Positive (reversal of anisocoria)	15
Negative	31
Observation time, min, mean (range)	89.1 (40-140)

Discussion

Apraclonidine is a sensitive and specific test for establishing a diagnosis of Horner syndrome in the pediatric population.¹¹ A 2006 study¹² reporting 10 patients (age range, 1.2-18 years) noted adverse reactions in 2 patients (20%), in the form of conjunctival hyperemia. This was not noted in any of our patients.

In infants, topical brimonidine 0.2%, another α-2 adrenoceptor agonist, has been reported to cause central nervous system depression, respiratory depression, and hypothermia because of its ability to cross the immature blood-brain barrier.¹³ Similar concerns regarding apraclonidine testing in infants, especially those under the age of 6 months, evolved after a 2007 report describing a 5-month-old who experienced periocular blanching and lethargy 1 hour after testing with apraclonidine eye drops. The child was taken to the emergency room and found to be hypoxic and bradycardic. She improved on oxygen mask and was discharged home 2 hours later. The same report described 4 other cases that were reported through the pediatric ophthalmology Internet discussion group, which at that point included 600 members, who were asked to complete a questionnaire regarding side effects with apraclonidine testing in infants with suspected Horner syndrome. Of those 4 patients, 3 had lethargy and 1 10-week-old experienced unresponsiveness without hypoxia or bradycardia. No intervention was described for any of the 4 cases.¹¹ This report recommended using caution with apraclonidine eye drops in infants or avoiding their use especially in those under the age of 6 months. Apraclonidine is commonly used for its intraocular pressure–lowering effects in pediatric glaucoma.¹⁴ A study evaluating the rate of adverse reactions due to apraclonidine use in children with glaucoma in 2009 described an 8% rate of adverse reactions, none of which were life threatening.¹⁵ This study included 75 children with an age range of 0.1 months to 17 years. Six patients experienced some form of adverse reaction: 2 developed topical allergy, 3 developed lethargy, and 2 had a decreased appetite. All of the adverse reactions subsided with the cessation of treatment and did not require further intervention. This report described a median number of apraclonidine dose exposures of 15 (range, 4-5745), with 80% of the patients receiving more than 10 applications of apraclonidine. The authors noted that some infants may be more susceptible than others to the central nervous system effects of apraclonidine, but concluded that use of apraclonidine in pediatric glaucoma was relatively safe, even in young infants; however, they recommended warning parents to monitor for serious side effects.

To our knowledge, our study of 46 patients is the largest study evaluating the safety of apraclonidine eye drop testing in children presenting with anisocoria suspicious for Horner syndrome. Most of our patients were tested in the outpatient clinic setting; only 3 were tested while inpatient. None of our patients experienced adverse reactions to the single exposure to apraclonidine eye drops (0.5%), including patients under 6 months of age (n = 17), with the youngest being 7 weeks. No children were excluded because of underlying neurologic, cardiac, or other medical conditions, which were not described as risk factors in previous reports. Based on our results, severe systemic adverse reactions to apraclonidine eye drop testing for anisocoria are rare in children, even in young infants. Some nonthreatening adverse reactions may be related to repeated exposure to apraclonidine as evident from studies cited earlier. Horner syndrome is a serious condition, and in some cases malignant tumors, such as neuroblastoma, may present with isolated Horner syndrome in the first year of life.⁹ This was demonstrated in one of our patients, who presented at the age of 3 months with isolated Horner syndrome. Early recognition and intervention may be lifesaving.³

Testing with apraclonidine has some limitation as the abnormal pupillary response occurs as a result of denervation hypersensitivity. This involves the upregulation of α1 receptors on the surface of the denervated pupil dilating muscle fibers. This takes time to develop and may not be present immediately after onset of sympathetic interruption. This suggests that if there is a high clinical suspicion of Horner syndrome, further workup should be pursued even in the setting of negative apraclonidine testing.

This study had some limitations. As a retrospective study, it relied on the documentation of side effects observed by the physician or caretakers. Some minor side effects, such as transient irritation, may not have been observed or reported. As in previous studies, ours also relied on the observation of adverse reactions, without any objective measurement of heart rate, respiratory rate, or oxygen saturation. All previous reported adverse reactions were clinically detectable without measurement of physiological parameters. A prospective study with objective monitoring for children undergoing apraclonidine testing would help confirm our findings.

Literature Search

The National Center for Biotechnology Information was searched on March 29, 2021, using the following terms: apraclonidine, anisocoria, Horner, and pediatric.