Infant presents with tearing, discharge from both eyes

Ptosis was observed on the right side with anisocoria.

Issue: February 25, 2016

ByTimothy Winter, DO

ByKristen E. Dunbar, MD

A 4-month-old previously healthy full-term infant presented to her pediatrician with tearing and discharge from both eyes. The mother noted this had been present since birth and did not note any associated symptoms. On exam, the pediatrician noted normal development and bilateral tearing. The infant was diagnosed with nasolacrimal duct obstruction. Warm compresses and nasolacrimal massage were recommended.

The patient was sent home; however, a week later she returned to the emergency department with right eye injection and mild edema around the right eye. She was playful on exam and had appropriate intake/output for her age. The emergency department prescribed polymyxin B-trimethoprim eye drops, and the patient was again discharged. A week later, she returned to the emergency department with bilateral ocular discharge and injection in both eyes. Again, she had no change in activity or appetite. She was prescribed erythromycin ointment. She was referred to our clinic for non-resolution of symptoms. The mother noted near complete improvement of symptoms; however, she noted that the right eye “looked smaller than normal.” The patient had no change in activity, appetite, or urination and stool output and had no fevers. Medical history was noncontributory.

Examination

The patient was noted to fix and follow objects in each eye separately. On pupillary examination, pupils were noted to be 1 mm on the right and 2.5 mm on the left in photopic conditions and 3 mm on the right and 6 mm on the left in scotopic conditions. She was also noted to have ptosis on the right, with a marginal reflex distance-1 of 3 mm on the right and 5 mm on the left (Figure 1). Pupils were briskly reactive, and there was no afferent pupillary defect noted. The rest of the anterior segment and posterior segment exam was within normal limits.

Figure 1. This photograph in photopic conditions shows ptosis of the right upper lid with anisocoria. Although difficult to see, the right pupil is approximately 1 mm in diameter and the left is approximately 2.5 mm. The second photo is in scotopic conditions and also shows ptosis, but it also shows a larger difference in scotopic conditions.

Before dilation, the options for testing the etiology of the anisocoria were discussed with the family. Given studies documenting extreme drowsiness, unresponsiveness, bradycardia, hypertension and decreased oxygen saturation in infants, the decision was made to defer Iopidine (apraclonidine, Alcon) testing. Topical cocaine testing was discussed with the mother; however, given the length of time necessary for compounding the agent at our institution, the mother declined testing with topical cocaine. Photographs of the child from 1 month prior were requested but did not show any ptosis or anisocoria at that time.

What is your diagnosis?

Anisocoria, unilateral ptosis

The differential diagnosis of Horner’s syndrome is broad because it can be caused by a disruption anywhere along the oculosympathetic pathway. This includes both congenital and acquired causes. The oculosympathetic pathway originates in the hypothalamus and descends through the midbrain, pons and medulla to the cervical spinal cord where it synapses in the spinal cord. From there, it courses through the stellate ganglion to synapse at the superior cervical ganglion. It then courses along the internal carotid artery through the cavernous sinus to innervate the orbit. An injury or mass at any point along this tract causes damage to the sympathetic nervous system, ultimately resulting in a Horner’s syndrome characterized by unilateral ptosis, miosis and anhidrosis.

PAGE BREAK

Acquired etiologies include carotid injury or dissection, injury to the brachial plexus, stroke, tumor, brainstem injury and apical lung tumors. In a young child, congenital etiologies are most common, most frequently caused by damage to the brachial plexus during delivery. We requested prior photos of the child, and it was noted that the ptosis and anisocoria were indeed present since birth.

Diagnosis and management

A discussion was raised as to what type of imaging was best to determine the etiology of the Horner’s syndrome; however, before the imaging could be scheduled, the next day the child was brought in to the emergency department after there was concern for non-accidental trauma. A skeletal survey was performed, and a mediastinal mass was incidentally seen on thoracic films (Figure 2). The child was admitted for further work-up including mediastinal mass biopsy, which returned positive for neuroblastoma. The child was started on a chemotherapy regimen of etoposide and carboplatin. Urine catecholamine testing (HVA/MVA) was negative.

Figure 2. Supine film showing right superior chest paraspinal mass (marked by arrow).

Discussion

Horner’s syndrome in a child can be a particularly challenging diagnosis to make, especially in fussy, uncooperative patients. As a result, it can easily be overlooked. Testing for Horner’s syndrome in the pediatric population can be equally challenging. Iopidine is contraindicated in infants, and cocaine, depending on the hospital, can be difficult to obtain and despite a lack of systemic side effects is often deferred by the patient’s parents given the association with recreational drug use.

As a result, diagnosis and determining the etiology of Horner’s syndrome are often left to imaging modalities. Specific guidelines for imaging the pediatric population have not been well-established. A retrospective review of 23 infants with Horner’s syndrome found that Horner’s syndrome presenting in the first year of life was usually a benign entity. In 70% of these cases, no cause was determined. This led to the author’s recommendation that in infants, if no cervical or abdominal masses were noted on exam and there was lack of involvement of other cranial nerves, only spot VMA should be checked. If VMA is found to be negative, routine follow-up with the pediatrician was the only further work-up warranted. Other studies concluded similar findings, while others felt that further imaging was necessary given the risk of uncovering a potentially life-threatening etiology.

The type of imaging in Horner’s syndrome in infants is another diagnostic challenge without a clear consensus. CT imaging with and without contrast is typically recommended given the high prevalence of calcifications in neuroblastoma because these calcification are more difficult to see on MRI. Despite this, one retrospective study by Mahoney recommended MRI for imaging Horner’s syndrome given an MRI’s lower exposure to superior resolution of structures.

References:
Al-Moosa A, et al. Curr Opin Ophthalmol. 2011;doi:10.1097/ICU.0b013e32834bbf06.
Bell RL, et al. J Trauma. 2001;doi:10.1097/00005373-200108000-00034.
Cahill JA, et al. J Emerg Med. 2015;doi:10.1016/j.jemermed.2014.07.041.
George ND, et al. Br J Ophthalmol. 1998;doi:10.1136/bjo.82.1.51.
Jeffery AR, et al. J AAPOS. 1998;doi:10.1016/S1091-8531(98)90008-8.
Mahoney NR, et al. Am J Ophthalmol. 2006;doi:10.1016/j.ajo.2006.05.047.
Roberts DR, et al. Brain Dev. 2015;doi:10.1016/j.braindev.2015.08.009.
Smith SJ, et al. Arch Ophthalmol. 2010;doi:10.1001/archophthalmol.2010.6.

For more information:
Kristen E. Dunbar, MD, and Timothy Winter, DO, can be reached at New England Eye Center, Tufts University School of Medicine, 750 Washington St., Box 450, Boston, MA 02111; website: www.neec.com.
Edited by Kristen E. Dunbar, MD, and Kendra Klein, MD. They can be reached at New England Eye Center, Tufts University School of Medicine, 750 Washington St., Box 450, Boston MA 02111; website: www.neec.com.