Question of the Week
Video Library
Ophthalmology Books & Manuals
Cyber-Sight Atlas of Eye Diseases
The Ophthalmology Minute
Nursing Education
Eye Care Equipment
ORBIS Program Features
Free Online Journals
Ophthalmology Links
Ask a Professor
Print ViewPrint this Page
Pupillary Abnormalities -  

Horner Syndrome
Lecture 12 of 13  NEXT»

Horner syndrome, which affects the eye and eyelids, is caused by paralysis of the cervical sympathetic nerves.

Anatomic Considerations

Discussion of the many syndromes involving the long sympathetic nerve chain to the eye requires at least a cursory knowledge of the relevant anatomy (Figs. 3.6, 3.7).

fig. 3.6

Figure 3.6 Course of sympathetic nerves from hypothalmus to the eye and adjacent anatomic structures.

fig. 3.7

Figure 3.7 Clinical differences of three sympathetic nerve.


PTOSIS. Ptosis is never severe in Horner syndrome, since the levator portion of the third cranial nerve does most of the lid elevation. The ptosis varies, depending on how tired or alert the patient is. It is not uncommon to observe a ptosis when the patient is evaluated after admission in the evening (when tired) and to find it has improved in the morning. This improvement is not real, since the levator and frontalis muscles and uninjured sympathetic fibers act to overcome the ptosis. Occasionally, no ptosis exists in the presence of sympathetic fiber damage, namely when the fiber to Müller's muscle has already branched off in the orbit proximal to the site of injury. Müller's muscle may also be spared in cases of slight damage to the sympathetic fibers. I feel that this sparing is rare. The lack of ptosis makes the diagnosis of Horner syndrome more difficult, but its absence or presence does not usually help much in determining the anatomic location, since the other factors just mentioned influence the degree of ptosis.

An additional problem in diagnosing the ptosis of Horner syndrome occurs in the elderly patient who may also have pseudoptosis secondary to blepharochalasis. In such a case, attention should be directed to the lower lid, which also has sympathetic innervation. I prefer to call this reaction Kearns' lower lid sign, since it was Kearns who pointed out to me the significance of sympathetic innervation of the smooth muscle of the lower lid. This innervation normally holds the lower lid down and against the globe. When the muscle is paralyzed, the lid rides up slightly (Fig. 3.8, A and B). The test for the lower lid sign is performed by having the patient fixate on a hand light and then moving the light up until the side with the suspected Horner syndrome has the 6 o'clock position of the cornea barely touching the lower lid. Then the other eye is observed. Instead of being in the same position, some white sclera shows between the lower cornea and the lower lid. Thus in Horner syndrome there is upper lid ptosis !Ind often lower lid elevation as well.

fig. 3.8a
fig. 3.8b

Figure 3.8 Hydroxyamphetamine test on third-neuron Horner syndrome. A. Before administration of the drug, the affected right pupil is smaller than unaffected left pupil. B. Thirty minutes after topical hydroxyamphetamine, the right pupil does not change, and the left pupil dilates. Notice that less sclera is seen between the lower lide and 6 o'clock position of limbus in the right eye thatn in the left. This elevation of the lower lid on the affected side is seen all types of Horner syndrome regardless of which neuron is involved.

APPARENT ENOPHTHALMOS. The position of the upper and lower lids makes the pitlpebral fissure narrower in Horner syndrome; thus less of the eye is seen. Unlike man, some animals have functional smooth muscle in the orbit, and paralysis of this muscle results in enophthalmos. In man, the position of the lids gives the impression of enophthalmos; however, the enophthalmos cannot be documented by exophthalmomeler readings. It is apparent rather than real.

MIOSIS. Miosis may seem to vary in Horner syndrome. The brightness of the background light in which the patient is examined may make the anisocoria difficult to detect. The same can be said of the physician standing in front of the patient during the examination and thereby stimulating the patient's accommodation and reducing the size of both pupils. To overcome these two problems, the patient should be examined in a semidark room with his or her gaze at a distant point and the observer off to one side. Because of the slowness of the affected pupil to dilate in the dark, the anisocorla is greater during the first 5 seconds of dark adaptation than at 15 seconds.

The miosis may also vary according to the extent of the defect, patient alertness, extent of reinnervation, and degree of denervation sensitivity. As a result of denervation supersensitivity, a patient may complain of an occasionally larger pupil on the side on which the physician observes a smaller pupil. This situation rarely occurs, but it probably represents the patient's supersensitivity to circu-lating epinephrine, a phenomenon that can be elicited by the instillation of one drop of a 1:1000 aqueous epinephrine solution topically onto both corneas. In the patient with Horner syndrome, the pupil that has developed supersensitivity dilates, whereas the other pupil shows no reaction.

Systemic drugs such as barbiturates and narcotics can also cause miosis, but without ptosis. A notable exception to the rule is glutethimide (Doriden), which tends to enlarge the pupils, particularly when taken in toxic doses. Topical medications (including most of the antiglaucoma drugs) also cause miosis.

Miosis is greater in postganglionic lesions than in preganglionic ones.

OCULAR HYPOTONY. The intraocular pressure on the side of a Horner syndrome is at least 5 mm less than the pressure in the fellow eye.

HETEROCHROMIA. The human iris is blue or slate gray at birth. Those irises that become brown do so by the end of the first year of life, a change that involves sympathetic innervation. Therefore, the iris of a patient with congenital or neonatal Horner syndrome usually remains lighter, since no sympathetic stimulation occurs to make it darker blue or brown than the iris of its fellow eye. This phenomenon is obvious if the other eye is brown and the eye with the Horner syndrome is blue. It is not as apparent in blue-eyed patients, in whom the difference may be more subtle and can easily be overlooked. The difference is more easily seen in daylight than in artificial indoor light. Depigmentation is said to occur in acquired adult Horner syndrome, but for all practical purposes it can be considered a rare phenomenon. Therefore, Horner syndrome with heterochromia has a congenital or neonatal onset—a fact that may be of value in medicolegal testimony (Fig. 3.9).

fig. 3.9

Figure 3.9 One example of congenital Horner's with heterochromia. Right eye has brown iris, and left iris is hazel with smaller pupil. Right iris is brown and left iris blue, but in this case the pupils were dilated prior to the picture. (Pictures courtesy of Dr. Caleb Gonzalez.)

INCREASE IN ACCOMMODATION. Cogan demonstrated that an increased amplitude of accommodation occurs on the side with Horner syndrome. He also proved that the increase is caused not by the miosis but by change in the ciliary muscle that amounts to 0.5 to 1.5 diopters. It is difficult to use the phenomenon as a test in patients under 35 years of age because young people can read almost up to their nose; thus any variations in accommodation are difficult to detect clinically. In examining someone who needs glasses, be sure that the patient is wearing the correct distance prescription, with an equal reading prescription over it. With the use of the near reading card, you will find that the patient reads significantly closer on the side with Horner syndrome.

ANHIDROSIS. Lesions from the posterior hypothalamus to the bifurcation of the carotid artery result in ipsilateral loss of sweating ability on the face. If Horner syndrome occurs above the carotid artery bifur-cation, the sweating mechanism is intact, a good localizing sign. However, demonstrating the loss of sweating ability or eliciting a history of it from the patient is not easy.

DIPLOPIA. Horner syndrome with contralateral diplopia suggests a trochlear nerve palsy on the opposite side that is found in a brainstem lesion. If the trochlear nerve paresis is on the same side as the Horner lesion, consider the cavernous sinus or superior orbital fissure as the anatomic location.

Chemical Tests

The clinical usefulness of cocaine and epinephrine was outlined by Foerster and Gagel. These authors believed that using either a 4% cocaine solution or a 1:1000 aqueous epinephrine solution can differentiate the neurons anatomically. Cocaine prevents the reuptake of norepinephrine at the motor endplate and thus prolongs its action on the effector cell. If the sympathetic pathways are interrupted, norepinephrine should not be released, and therefore, a mydriatic effect should not occur. Epinephrine, on the other hand, works directly as a stimulator of the motor endplate. A 1:1000 aqueous epinephrine solution should not dilate a normal pupil; however, if supersensitivity exists (as can occur in postganglionic third-neuron lesions), the pupil will dilate. Therefore, Foerster and lagel believed that the scheme shown in Figure 3.10 was a good one for localizing a Horner syndrome to one of the three sympathetic neurons; however, the scheme usually does not work.

fig. 3.10

Figure 3.10 Effects of cocaine and epinephrine on the pupils in Horner's syndrome (Forster and Gagel schema).

Third-neuron Horner syndrome rarely shows supersensitivity to 1:1000 aqueous epinephrine. It is better to use II or 2% phenylephrine. This drug may dilate both pupils but will have a greater effect on the supersensitive pupil. It is important to document both pupil signs before and after the test to evaluate the relative effectiveness of the drug on both pupils. In theory, cocaine should not dilate any Horner syndrome pupil, no matter what neuron is involved, since no impulse comes to the motor end-plate to release norepinephrine in the first place. If the Horner syndrome pupil does dilate, it is probably because the Horner syndrome is incomplete or because a small amount of norepinephrine has been released at the endplate constantly without direct central nervous system stimulation. Thompson and Mensher have modified Foerster and Gagel's scheme by the use of hydroxyamphetamine, which works differently from the other two agents by releasing endogenous norepinephrine from an intact motor endplate. If the condition is a third-neuron Horner syndrome and the nerve and endplate have degenerated, epinephrine Is not present and thus no mydriasis occurs with hydroxyamphetamine.

If the condition involves a first or second neuron, the third neuron is left intact, with the norepinephrine stores present. Even though no central nervous system innervation exists, hydroxyamphetamine releases the norepinephrine, and the pupil should dilate. My experience with the hydroxyamphetamine test makes it seem valuable, I think that it has good theoretic points to recommend it and that it should be tried. Maloney, Younge, and Moyer evaluated the hydroxyamphetamine test and found it 84% accurate in predicting and 96% accurate in confirming a third-neuron Horner syndrome. Grimson and Thompson agreed that the test may be inconclusive in 15 to 20% of cases, and this number rises unless photographs are taken. The test is particularly unreliable in cases of congenital Horner syndrome. My experience with the test has been more limited than that of these authors, but I find it more useful than cocaine and 1:1000 aqueous epinephrine.

The cocaine test should be discussed further. The cocaine should be in a 4% solution and not a 10% solution as some advocate. A 10% solution adds nothing to the test, and copious use may do some transient damage to the epithelium. Changing the corneal tear film or epithelium by drops or corneal sensitivity testing only alters the topical pharmacologic sensitivity tests such as epinephrine and mecholyl and hydroxyamphetamine.

Causes and Significance

A usual question is, Why bother about Horner syndrome since most patients are not symptomatic or, at the worst, have only a slight ptosis? The reason for bothering is that Horner syndrome may be the tip of an iceberg, indicating a more serious condition.

In determining the cause of Horner syndrome, the usual approach is to try to localize it to a specific neuron. In older patients, one of the most common causes of Horner syndrome is a vascular infarct of the sympathetic chain. The lesion may occur along the first neuron in the brainstem, owing to the obstruction of the small penetrating vessels from the basilar artery. A lesion also may occur along the distribution of the third-neuron sympathetic chain associated with the carotid artery. The latter has been shown by Sears, Kier, and Chavis to occur experimentally.

The congenital variety of Horner syndrome is considered by most to be caused by neck injury from manipulation during a difficult forceps or breech delivery. Trauma is the most frequently found cause in those beyond infancy and under the age of 21. Patients older than 20 years of age, and particularly those over 50 years of age, with second-neuron Horner syndrome, should be investigated for the presence of tumor (usually malignant) when the onset of an isolated Horner syndrome occurs. The most common tumors are metastatic and bronchogenic carcinoma, particularly apical, or Pancoast tumors. Benign tumors, such as neurofibromas and thyroid adenomas, are less common.

The causes of isolated Horner syndrome have changed somewhat since the report of Giles and Henderson. Third-neuron Horner syndrome carries a high incidence of nontumor disease as the cause. Third-neuron Horner syndrome is more likely caused by a headache syndrome (e.g., Raeder's, cluster), trauma (e.g., basal skull fracture), or inflammations (e.g., Tolosa-Hunt, otitis media, herpes zoster, epidural anesthesia, and carotid artery dissection).

I have seen Horner syndrome in dissection of the internal carotid artery. Several of these have been from chiropractic manipulation. Carotid angiography can usually identify the problem. Some may be missed by this imaging technique such as in a case reported by Brown et al. They identified the dissection by MRI in which the subject demonstrated an area of hyperintensity.

The second-neuron Horner syndrome still carries a high incidence of tumors as the cause, as was pointed out by Giles and Henderson and confirmed in the subsequent reviews by Grimson and Thompson and by Maloney, Younge, and Moyer. Grimson and Thompson stated that the incidence is as high as 50%, and Maloney, Younge, and Moyer found that 72% of the tumors causing a Horner syndrome were along the second-neuron distribution. Younge makes an additional observation: Most of the tumors in their series were Pancoast type, and Horner syndrome was rarely the presenting sign. Their patients also had arm pain, which should serve as a clue to the real cause of the Horner syndrome.


How does the physician evaluate Horner syndrome once the diagnosis has been made? The time of onset may be difficult to establish. The time the patient says it started may only be the time he or she noticed it. One of the best ways to establish the time of onset is by examining old photographs of he patient, particularly job, army, or passport identification photographs. Such photographs are usually unretouched, and often the pupils can be seen with a magnifying glass or the large indirect ophthalmoscope lens. If these types of photographs are not available, ask the patient to supply some high school or college yearbook pictures or some wedding pictures. The fact that a Horner syndrome has been present for 10 years or more points to a benign cause, even when the exact reason for the condition remains obscure.

Ask the patient about any neck operation (e.g., a thyroid surgical procedure) that might have injured the sympathetic chain. Scars from such operations can easily be overlooked. Ask also about a chest or heart operation, which could also account for Horner syndrome. In the past, when more carotid arteriograms were done using the direct carotid artery injection route, both permanent and transient Horner syndromes were sometimes seen.

Question the patient carefully about any neck trauma as a youth. If the patient played sports, ask whether a neck collar was worn for several months because of an athletic injury. Photographs showing the patient's pupils around that time, when any miosis would have been more evident before reinnervation, may document the time of onset. Horner syndrome can also be seen in severe whiplash injuries, such as those occurring in automobile accidents.

The examination should include palpation of the neck for masses, thyroid nodules, or supraclavicular nodes. Apical views of the lung to visualize a Pancoast tumor, as well as skull and cervical roentgenograms, should be ordered. Hematologic evaluation to rule out lymphomas or Hodgkin's disease is also indicated.


FIRST-NEURON HORNER SYNDROME. The most common cause of first-neuron Horner syndrome is vertebral-basilar insufficiency as in Wallenberg syndrome. A Horner syndrome may be the only residual sign of a transient ischemic episode. Severe osteoarthritis of the neck, with obvious bony spurs visualized on roentgenographic examination, is known to have caused compression of the sympathetic fibers as they leave the cervical canal. I have also seen Horner syndrome occur with severe whiplash Injury without other obvious neurologic deficits. Horner syndrome can be transient or permanent, and it may be the only evidence of the severity of the injury, particularly when the patient has multiple posttraumatic complaints.

Neurologic signs that appear to he on opposite sides of the nervous system suggest multiple sites of origin. One such combination is a Horner syndrome on one side and a fourth nerve palsy on the other side. However we must remember that the fourth nerve is the only crossed cranial nerve. The left fourth nerve is on the right side before it crosses in the anterior medullary velum. Therefore, this left fourth nerve before it crosses is near the right sympathetic nerve and can give this crossed arrangement of signs from one lesion.

SECOND-NEURON HORNER SYNDROME. Among the most common causes of second-neuron Horner syndrome is apical lung cancer (such as Pancoast tumor), which is best demonstrated by apical views of the lung. Mediastinal tumor is also a cause, but it is not as easily detected.

PHRENIC NERVE SYNDROME. A new triad involving a Horner syndrome has been reported: Horner syndrome, a hoarse voice, and paralysis of the hemidiaphragm. This triad involves the second neuron of the sympathetic chain, the phrenic nerve, and paralysis of the recurrent laryngeal nerve. The only place these three nerves are in close apposition is at the level of the sixth cervical vertebra. This syndrome occurs with local or recurrent tumors to this area. The three cases reported by Rowland Payne involved recurrences of breast carcinoma.

THIRD-NEURON HORNER SYNDROME—GROUP ONE. Third-neuron Horner syndrome can be broken down into two groups. The first group includes patients with Raeder's paratrigeminal syndrome, cluster headaches, and migraine. These three conditions may be aspects of the one disease, but each condition is worthy of comment because of its individual characteristics.

Raeder's Paratrigeminal Syndrome.
Raeder's paratrigeminal syndrome is essentially a painful Horner syndrome; the major complaint is pain over the first and second divisions of the trigeminal nerve. The patient shows all the signs of Horner syndrome, except that the sweating mechanism is intact because the lesion is located above the bifurcation of the carotid artery. There may be a small area of anhidrosis on the ipsilateral forehead. This area is supplied by terminal frontal branches coming in with the carotid artery. This is a variable sign. Raeder's syndrome occurs overwhelmingly in men, usually in those in their forties. The major consideration is that the condition is benign. The pain subsides, but frequently the Horner syndrome remains. The first reported case of Raeder's syndrome involved a meningiorna of the gasserian ganglion, but most subsequent reported cases have been benign. Davis, Daroff, and Hoyt reported one case associated with an extracranial aneurysm. Law and Nelson also had one case of a supraclinoid aneurysm, and Cohen, Zakov, and Solanga reported two cases associated with fibrous dysplasia.

Raeder's syndrome can be broken down into two groups. Group 1 exhibits Horner syndrome, trigeminal pain, and other parasellar cranial nerve involvement and is not benign. Group 2 exhibits the first two signs but lacks parasellar nerve involvement. Group 2 is usually caused by a vascular or inflammatory mechanism affecting the sympathetic chain and sensory fibers that run in the carotid artery sheath and is benign. If the pain is atypical or persists more than 3 months, these patients should be investigated even in the absence of other cranial nerve involvement. No specific treatment exists.

Cluster Headaches. Patients with third-neuron Horner syndrome may have headache as a primary finding. These patients are also usually in their forties. The headache is not steady, and it is predictable as to time of day and duration (usually several hours). These patients also have tearing and ocular hyperemia, as well as ipsilateral nasal stuffiness. The associated Horner syndrome may be transient. The condition is benign. At one time it was treated with histamine desensitization, but this treatment has fallen out of favor. In any case, it is usually a self-limited process, although episodes do recur (hence the name cluster headaches).

Migraine. In any form, even with a typical homonymous scotoma, migraine may affect the carotid artery and result in Horner syndrome. The condition is usually self-limited.

THIRD-NEURON HORNER SYNDROME—GROUP TWO. The second group of third-neuron Horner syndrome involves the facial sweating mechanism. The syndromes are (a) idiopathic hemifacial hyper-hidrosis, (b) postsympathectomy facial hyperhidrosis, and (c) hemifacial anhidrosis.

Idiopathic Hemifacial Hyperhidrosis. The patient with idiopathic hemifacial hyperhidrosis syndrome complains only of increased sweating on one side of the face, particularly on the forehead. Usually, the condition is aggravated by eating spicy foods. No other signs of Horner syndrome are present. The increased sweating is caused by the overactivity of the sympathetic fibers that subserve sweating, and it tllnnppears with sympathetic nerve block-air. This condition probably results from aberrant regeneration of the sudomotor fibers along the external carotid.

Postsympathectomy Facial Hyperlildrosis. After a complete cervical sympathectomy, anhidrosis is present in the ipsilateral affected area. Then after a long time, adjacent autonomic fibers from the vagus nerve sprout to innervate the sympathetic nerves. Thereafter, patients sweat and tingle when they eat. Since sweating is cholinerple and tingling is caused by pilomotion, which is adrenergic, the vagus nerve must make preganglionic connections where all fibers are of the cholinergic variety. No associated pupillary or palpebral signs are present.

Hemifacial Anhidrosis. Sweating over the ipsilateral face is lost with sympathetic interruption below the bifurcation of the carotid artery, where the facial sweat fibers leave the artery. Occasionally, however, sweating may be preserved on the forehead because some sympathetic fibers have gone with the orbital division, innervating the forehead directly. Therefore, anhidrosis of the face below the eye and above the upper lip has the same significance, even if the sweating mechanism remains intact in regard to the forehead.

Just as destructive lesions of the sympathetic chain cause a Horner syndrome, there are rare cases in which those same lesions can stimulate the sympathetic nerve. This produces mydriases either intermittently or permanently. Less frequently, there can also be lid retraction rather than ptosis, and more rarely, hyperhydrosis. This syndrome has been reported with malignant lung tumors and cervical cord disease. In the few cases that I have seen, this finding of stimulation is transient and becomes a destructive lesion with the signs of a Hornmer syndrome in a very brief period of time.

Another form of intermittent pupil dilatation is that in a patient who had a third-neuron Horner syndrome and now has supersensitivity to the circulating effective substance. Supersensitivity can develop as early as 2 to 3 weeks after sympathetic paralysis. The pupil supersensitivity will decrease as the sympathetic axons regenerate or sooner if there is reinervation by lateral sprouting of axons, which is most common in the autonomic system.

The first case of this I noted was in my own secretary. She arrived at the office one Monday morning with a severe headache, minimal left hypertropia, about a 1 mm larger left pupil, and ptosis. I diagnosed a third nerve palsy and arranged for the neurosurgeons to evaluate her. They did this and confirmed my diagnosis of an aneurysm. They operated successfully, and she had a full asymptomatic recovery. She was aware that an enlarged pupil was a sign of a third nerve palsy and an aneurysm. A few months later she noted intermittent dilatation of that same pupil without any other signs or symptoms. She was concerned that the aneurysm clip was slipping and that her aneurysm was again causing symptoms. The neurosurgeons and I worked her up and reassured her this was not the case. When we explored the history further, she noted that she would occasionally notice this pupillary difference while examining herself in the mirror. We surmised that she became anxious and produced more adrenaline, which then further dilated her pupil. She tested positive to 1/1000 aqueous adrenaline. Our conclusion was that at the time of her aneurysm compressing the third nerve, there must have been sympathetic injury that was masked by the mydriasis and ptosis of the partial third nerve palsy. She improved and has had no further problems.





Lecture 12 of 13 «Previous Lecture   1 2 3 4 5 6 7 8 9 10 11 12 13    Next»
footer logo