 |
Slide 1 |
Andrew G. Lee, MD
Introduction
Horner’s syndrome is an important clinical marker of potentially serious structural disease and results from disruption of the ocular sympathetic pathway. The three-neuron arc begins in the hypothalamus (first-order neuron), descends the posterolateral brainstem to the spine (C8-T2), exits as the second-order neuron, travels over the lung apex (eg, Pancoast tumor), travels up the neck along the cervical sympathetic chain, and synapses in the superior cervical ganglion. The third-order neuron travels on the internal carotid artery into the cavernous sinus, for a short course on the sixth cranial nerve, and with the trigeminal nerve (V1) into the orbit through the superior orbital fissure. Ophthalmologists should recognize the clinical features of Horner’s syndrome as the underlying lesion may be life-threatening (eg, carotid artery dissection or apical lung tumor).
|
The classic clinical features of Horner’s syndrome are ipsilateral ptosis, miosis, and anhidrosis. Although the upper lid ptosis is typically obvious, lower lid ptosis ("upside down ptosis") may also occur. The upper lid ptosis is usually mild (1 mm to 2 mm) due to involvement of the sympathetically innervated Mueller’s muscle in the lid and thus complete ptosis (eg, as seen in third-nerve palsy) does not occur. The combination of the upper and lower lid ptosis produces the appearance of a narrow palpebral fissure (ie, apparent but not true enophthalmos).1-4
The anisocoria in Horner’s syndrome is generally mild and may be as little as 1 mm. The miotic (smaller pupil) pupil results from a lack of sympathetic innervation to the iris dilator muscle and, therefore, the anisocoria is greatest in darkness (Slide 1 and Slide 2). In ambient light, the anisocoria may be minimal and it is important to check for anisocoria in room light, dark, and with added light. In addition, the normal pupil dilates rapidly (within 5 seconds) after the light is turned off. In contrast, the pupil in the Horner’s syndrome dilates more slowly over about 15 seconds to 20 seconds (ie, "dilation lag"). I use clinical photographs (Polaroid) of the pupil to document the anisocoria in light and dark, to demonstrate dilation lag at 5 seconds vs. 15 seconds, and to compare the anisocoria at baseline and following pharmacologic testing.
|
The sympathetic plexus also supplies sweating to the ipsilateral body and face. Patients with central (first-order neuron) or preganglionic (second-order neuron) lesions may have loss of sweating (ie, anhidrosis) on that side of the hemibody. Lesions at the junction of the internal and external carotid artery bifurcation may produce sweating abnormalities on the ipsilateral face. The post-ganglionic sympathetics travel with the internal carotid artery and innervate sweat glands only to the medial forehead. It has been my experience that anhidrosis is an uncommon complaint even with pharmacologically confirmed Horner’s syndrome. Although starch testing to localize anhidrosis has been reported previously, I find it to be of limited clinical value. Iris heterochromia may occur in Horner’s syndrome (lighter colored eye) because the sympathetic innervation is involved in the development of iris pigmentation. Although this finding is suggestive of a congenital Horner’s syndrome, it may occur in acquired cases.
Central (first-order neuron) Horner’s syndrome is typically associated with other neurologic signs and symptoms (eg, Wallenberg lateral medullary stroke syndrome). It is uncommon for ophthalmologists to be the first or only provider to evaluate a central Horner’s syndrome. On the other hand, a preganglionic second-order neuron Horner’s syndrome may present in isolation and still be due to a serious underlying etiology (eg, neoplasm of the neck or pulmonary apex—the Pancoast tumor, or a carotid dissection). Although many isolated cases of postganglionic Horner’s syndrome are benign, patients may also have an underlying carotid dissection. I recommend imaging for all pre-ganglionic Horner syndromes and consideration for imaging post-ganglionic Horner’s syndrome, especially when associated with severe pain.
Pharmacologic Testing
Typical Horner’s syndrome findings (eg, ptosis, upside down ptosis, miosis, and dilation lag) are sufficient to make the diagnosis of Horner’s syndrome versus physiologic anisocoria. Pharmacologic testing, however, will differentiate between physiologic anisocoria and Horner’s syndrome.5-12
|
Cocaine inhibits reuptake of norepinephrine at the neuromuscular junction. Topical cocaine (4% to 10%) will dilate the normal pupil. In patients with Horner’s syndrome, however, because there is less or no norepinephrine at the postganglionic junction, there will be relatively less (or no) dilation to topical cocaine. A post-cocaine anisocoria of greater than 0.8 mm is considered diagnostic of Horner’s syndrome (Slide 3). Dilation of the suspected miotic pupil alone is not sufficient to exclude Horner’s syndrome and may occur in partially denervated Horner’s syndrome patients. Thus, pre- and post-cocaine anisocoria should be compared directly. In contrast to cocaine, hydroxyamphetamine releases norepinephrine from the postganglionic neuron and, if the lesion involves the third-order neuron, then the involved pupil will not dilate as well as the fellow eye. Preganglionic Horner’s syndrome is suspected if the hydroxyamphetamine testing dilates the pupils and resolves the anisocoria (after a positive cocaine test). Patients undergoing pharmacologic testing should be advised regarding the possibility of producing a false positive drug screen for illicit substances if tested at or around the time of topical testing for Horner’s syndrome.
The pharmacologic testing does not differentiate between the first- and second-order preganglionic Horner’s syndrome. I perform the cocaine test first to confirm Horner’s syndrome and then localize with the hydroxyamphetamine (on a different day). Patients with congenital Horner’s syndrome may not have reliable hydroxyamphetamine testing because of possible orthograde transsynaptic dysgenesis of the postganglionic neuron. I, therefore, recommend confirming the Horner’s syndrome with cocaine and the proceeding to neuroimaging of the entire oculosympathetic pathway in unexplained cases in childhood.
Some miotic pupils are due to myogenic rather than neurogenic (ie, Horner’s syndrome) etiologies. Topical 1% phenylephrine, a direct sympathomimetic, will dilate a normal dilator muscle of the iris. If the pupil does not dilate, this suggests that the dilator muscle rather than the nerve alone is responsible for the miosis, which is important because the "little old Adie’s" tonic pupil or other abnormality of the dilator muscle will not dilate to topical cocaine and will produce a false-positive test.
Apraclonidine (Iopidine; Alcon Laboratories, Fort Worth, Tx), which is an alpha-receptor agonist, has also been used to confirm Horner’s syndrome. Apraclonidine causes reversal of the anisocoria in Horner’s syndrome with the abnormal miotic pupil becoming larger due to denervation hypersensitivity 1-receptors in the pupil dilator muscle. There is also reversal of the ptosis and blanching of the conjunctiva. Apraclonidine testing may eventually prove to be superior to cocaine testing in Horner’s syndrome because the drug was originally developed as a glaucoma agent and, therefore, corneal permeability is not an issue. In addition, compared to topical cocaine and hydroxyamphetamine, apraclonidine is more readily available.10-12
Evaluation
If the cocaine test is negative, I assume that the patient has a physiologic anisocoria. No imaging or further evaluation is generally required. If, on the other hand, the cocaine test is positive for Horner’s syndrome, then I recommend investigation for an underlying etiology. I proceed to localization with hydroxyamphetamine to direct the imaging studies appropriately. In patients with acute Horner’s syndrome and suspicion for an underlying emergent condition (eg, carotid dissection) I image the entire oculosympathetic axis (ie, MRI of head and neck to T2 in the chest with neck MRA). If Horner’s syndrome is not emergent, then I schedule the patient for imaging of the entire oculosympathetic axis with radiology but tailor the imaging study to the pre- or post-ganglionic pathway based on a hydroxyamphetamine test prior to the MRI.
There are multiple etiologies for Horner’s syndrome that include structural lesions anywhere along the sympathetic axis.13-56
Adults with a clear history of traumatic or surgical damage to the oculosympathetic pathway and children with a clear history of birth trauma and isolated Horner’s syndrome may not require imaging. Horner’s syndrome may occur after any surgery of the neck or chest, from placement of vascular access lines (eg, internal jugular or subclavian). Children with an unexplained Horner’s syndrome should be evaluated for occult neuroblastoma with imaging of the entire sympathetic axis.
Conclusion
Ophthalmologists should be aware of the clinical presentation of Horner’s syndrome. Differentiation from physiologic anisocoria can normally be done clinically, but some patients require pharmacologic confirmation with topical cocaine. I recommend pharmacologic confirmation with cocaine and localization with hydroxyamphetamine in adult patients. Iatrogenic and traumatic etiologies should be sought and review of prior old photographs may be useful to establish the duration of the Horner’s syndrome. Unexplained Horner’s syndrome should undergo imaging directed at the topical localization to the pre- or post-ganglionic Horner’s syndrome pathway. Children may have transsynaptic degeneration and, therefore, hydroxyamphetamine testing after cocaine confirmation of the Horner’s syndrome may not be reliable. Imaging the entire sympathetic axis may be necessary in children to exclude compressive lesion (ie, neuroblastoma).
References
