Neurotrophic corneal endothelial failure complicating acute Horner syndrome

Neurotrophic corneal endothelial failure complicating acute Horner syndrome

Neurotrophic Corneal Endothelial Failure Complicating Acute Horner Syndrome Ehud Zamir, MD, Itay Chowers, MD, Eyal Banin, MD, PhD, Joseph Frucht–Pery,...

361KB Sizes 0 Downloads 25 Views

Neurotrophic Corneal Endothelial Failure Complicating Acute Horner Syndrome Ehud Zamir, MD, Itay Chowers, MD, Eyal Banin, MD, PhD, Joseph Frucht–Pery, MD Purpose: The authors report the clinical findings of a unique case of rapid corneal endothelial decompensation in association with acute Horner syndrome. Study Design: Case report and literature review. Methods: The authors followed a 38-year-old woman who developed Horner syndrome after right jugular vein catheterization during cardiac valvular surgery. Shortly after the operation, Horner syndrome accompanied by conjunctival hyperemia and stromal corneal edema developed in the right eye. Over the course of 4 months, the eye became painful, the corneal endothelial cell count dropped precipitously, and the stromal edema worsened, causing a difference of 100 ␮m in central corneal thickness compared to the unaffected eye. Deep stromal vascularization started at the limbus, resembling interstitial keratitis. Results: A 3-week course of topical steroid treatment resulted in a dramatic improvement in the stromal corneal edema and regression of the deep stromal vascularization. Ocular and right hemicranial pain subsided shortly thereafter. Conclusion: The authors hypothesize that corneal endothelial failure in this unique case may have resulted from traumatic sympathectomy. According to experimental evidence in the reviewed ophthalmologic literature, sympathetic innervation may have a neurotrophic role in the cornea. Corneal pathology similar to the authors’ case has been described in hemifacial atrophy (Parry–Robson syndrome), a disorder that is assumed to result from sympathetic denervation and that can be produced in animals by cervical sympathectomy. The authors therefore hypothesize that sympathetic denervation of the cornea may rarely cause endothelial decompensation and corneal edema. To the authors’ knowledge, this is the first reported case of corneal endothelial failure in Horner syndrome. Ophthalmology 1999;106:1692–1696 Horner syndrome is best known for its conspicuous effects on lid position and dilator muscle tone, both representing interruption of sympathetic innervation, with resulting smooth muscle paresis. Facial skin anhidrosis may also appear, depending on the location of the sympathetic nerve lesion. Less-noticeable manifestations may include conjunctival hyperemia1 and ocular hypotony.2 Despite the fact that the cornea is dually innervated, by sensory as well as sympathetic nerve fibers,3 it is not known to be clinically affected by sympathetic denervation. In this article, we present a case of unilateral, primary corneal endothelial failure occurring in a patient with Horner syndrome. We review the experimental data supporting physiologic trophic effects of sympathetic innervation on the cornea and try to establish the relevance of these data to our case. To our knowledge, this case is the first clinical report of corneal endothelial failure in Horner syndrome.

Originally received: January 5, 1999. Revision accepted: May 27, 1999. Manuscript no. 99006. From Hadassah University Hospital, Jerusalem, Israel. Address correspondence to Ehud Zamir, MD, Department of Ophthalmology, Hadassah University Hospital, P.O. Box 12000, 91120 Jerusalem, Israel. E-mail: [email protected].

1692

Case Report A 38-year-old woman with rheumatic heart disease underwent aortic valvuloplasty for critical aortic stenosis. The patient’s ocular history was unremarkable, and her visual acuity, as recorded by her local ophthalmologist, had been 6/7.5 in both eyes. During the operation, a central venous catheter was inserted percutaneously into the right internal jugular vein. Immediately after the operation, anisocoria was noted. Bedside ophthalmic examination revealed 2-mm ptosis of the right upper lid and a “reverse ptosis” of the right lower eyelid. The right conjunctiva was mildly hyperemic. The right pupil was round and reactive, with a diameter of 4 mm, compared to the left pupil, which was 7 mm. The anisocoria increased in the dark. Five days later, the best-corrected visual acuity was 6/9 in the right eye and 6/7.5 in the left eye. Slit-lamp examination results showed sectoral perilimbal conjunctival injection in the right eye (Fig 1). The inferior and central stroma of the right cornea was edematous in a segmental distribution (Fig 2). There were no endothelial guttata in either eye. The corneal epithelium appeared normal and did not stain with fluorescein or rose– bengal. The anterior chambers were deep and quiet, and the intraocular pressure was 15 mmHg in both eyes. The lens, vitreous, and fundus were unremarkable in both eyes. A 4% cocaine test confirmed Horner syndrome in the right eye. Hydroxyamphetamine was not available for testing. However, there was no pupillary response to phenylephrine 0.5%. A preganglionic lesion was therefore suspected. Central corneal thickness, as measured by an ultrasonic pachymeter (DJH 2000; DJH Technology Inc., Frazer, PA), was 630 ␮m in the right eye and 530 ␮m in the left eye. Computerized specular microscopy (Konan Noncon Specular Microscope “Robo-CA”; Konan Inc., Nishinomiya,

Zamir et al 䡠 Corneal Edema in Horner Syndrome

Figure 1. Right Horner syndrome with ptosis and miosis. The photograph was taken in dim light; therefore, anisocoria is maximal. Note sectoral limbal conjunctival injection at the 4-o’clock and at the 7-o’clock positions. Lower lid shows a mild “reverse ptosis.”

Hyogo, Japan) showed normal endothelial morphologic analysis in both eyes, with cell density of 2600 cells/mm2 in the right eye and 2000 cells/mm2 in the left eye. During the following 2 months, the patient reported persistent ocular and right periorbital pain, as well as blurred vision. Visual acuity decreased from 6/9 to 6/12 in the right eye. The segmental stromal edema deteriorated progressively, with deep stromal vascularization starting at the inferotemporal limbus and penetrating centripetally to a distance of 2 to 3 mm from the limbus (Fig 3). The corneal nerves within the right stroma became considerably more prominent than those in the left

Figure 3. The temporal limbus of the right cornea. Note marked conjunctival hyperemia and deep stromal vascularization penetrating centripetally from the limbus at the 7- to 8-o’clock positions (arrows). These vessels regressed after topical steroid therapy.

Figure 2. Slit-lamp photograph of the right cornea. Stromal edema and “ground glass” appearance of the central cornea are shown.

cornea, which were barely visible (Fig 4), and the endothelium developed a “beaten metal” appearance. The endothelial cell count in the center of the right cornea dropped from 2600 to 900 cells/mm2 within 2 months, and the endothelial cells became markedly larger than those in the left cornea (Fig 5). The anterior chamber remained quiet. Serum Venereal Disease Research Laboratory results, skin tuberculin test, and human immunodeficiency virus serologic analysis were negative. There was no history of herpetic or other eye disease or varicella zoster. A trial of topical 0.5% methylprednisolone drops every 2 hours was started. During the following 4 weeks, the patient reported subjective alleviation of pain and improving vision in her right eye. The stromal edema cleared significantly, and the deep stromal blood vessels regressed

Figure 4. Thickened corneal nerves in the right cornea (small arrowheads). The stroma is thickened, and the endothelium has a beaten metal appearance (large arrowhead).

1693

Ophthalmology Volume 106, Number 9, September 1999

Figure 5. A, specular microscopy shows reduction of endothelial cell count in the right eye to 915 cells/mm2. Note enlargement and polymegathism of cells in the right eye compared with the normal mosaic in the left eye (B).

completely. Visual acuity improved to 6/9 in the right eye. Steroid therapy was discontinued without clinical evidence of relapse of the symptoms or signs. Four months after initial presentation, the patient’s corneal condition continued to be stable without any ocular treatment. The ultrasonic pachymetry of the central cornea decreased to 580 ␮m, and the endothelial cell count remained approximately 900 cells/mm2.

Discussion Our patient suffered an acute sympathetic neural insult that resulted in neurotrophic corneal changes. This interpretation is based on the simultaneous appearance of both Horner syndrome and keratopathy. We found additional support of this concept in the relevant clinical and experimental literature as reviewed below. The mammalian cornea is innervated by both sensory trigeminal fibers and sympathetic nerves originating from the superior cervical ganglion.3,4 There is evidence to suggest that sympathetic innervation affects the corneal hydra-

1694

tion status and other cellular functions by influencing both epithelial and endothelial physiology. Animal studies have established that corneal epithelial proliferation and wound healing are encouraged by intact sympathetic innervation.5–7 Corneal epithelial chloride anion transport, which has a partial role in stromal deturgescence, was shown in animals to be facilitated by adrenergic stimulation and increase in intracellular cyclic-adenosine monophosphate.8,9 Adrenergic receptors also are found in the corneal endothelium, where their activation increases intracellular cyclicadenosine monophosphate levels. This, in turn, affects cellular morphology, growth, and membrane permeability and promotes corneal deturgescence.10 –14 Despite extensive evidence from animal studies, the existence of sympathetic nerves in the adult human cornea had been disputed for years before finally being demonstrated by Toivanen and colleagues in 1987.14 The work of Marfurt and Ellis on human corneas15 indicated later that there is dense sympathetic innervation to the corneoscleral limbus and modest innervation to the cornea proper. However, there are only a few clinical studies that examined corneal changes following sympathetic denervation. Nielsen2 studied central corneal thickness in 14 patients with chronic Horner syndrome 6 or more months after onset. A small but statistically significant increase of 4.9 ␮m was detected between the mean thickness in the affected and normal contralateral eyes. However, none of the patients had clinical corneal edema or corneal symptoms. Although there are few data about corneal changes in isolated Horner syndrome, there are several reports of pathologic corneal changes in a closely related condition, the Parry–Robson syndrome (hemifacial atrophy). This is a chronic, neurotrophic disease thought by many to result from disruption of sympathetic denervation. Ocular involvement in this syndrome may include Horner syndrome, Fuchs’ heterochromic cyclitis, and iris heterochromia.16 –20 Cervical sympathectomy alone is sufficient to produce a similar syndrome that serves as a model of hemifacial atrophy in animals, including “keratitis and corneal ulceration.”21 Cornea guttata and bullous keratopathy were reported in a patient with hemifacial atrophy and Horner syndrome. Ford and colleagues22 recently described a similar case of unilateral, primary endothelial failure in a patient with hemifacial atrophy23 and suggested that perhaps a genetic defect in neural crest-derived tissues could underlie both abnormalities. We hypothesize that the common pathogenic factor contributing to corneal edema in our case of Horner syndrome and in Parry–Robson syndrome is sympathetic dysfunction. It is notable that the clinical picture in our case was characterized by chronic pain and hyperesthesia, along with morphologic changes in the corneal nerves. Chronic pain syndrome may accompany Horner syndrome.24 It is interesting in this respect that there exists experimental evidence for reciprocal trophic influence between sensory and sympathetic innervation in the cornea. Surgical sympathetectomy in animals leads to a marked increase in the number of sensory nerve fibers in the cornea and iris, and, conversely, sensory denervation leads to an increase in corneal sympathetic fibers.25–28 The normal neurotrophic effects on the

Zamir et al 䡠 Corneal Edema in Horner Syndrome cornea seem to depend on a balance in neuronal activity between sympathetic neurons and trigeminal sensory neurons.29 The enlargement of corneal nerves in the affected cornea of our patient could thus represent a reactive increase in sensory fibers, which may also account for the predominant ocular pain and hyperesthesia. We believe that our case, as well as the quoted literature, implies that sympathetic denervation may have neurotrophic effects on the human cornea. This sympathetic neurotrophic keratopathy is characterized by endothelial cell dysfunction and loss and by inflammatory signs. In our case, ocular pain predominated, together with sectoral conjunctival injection, deep segmental corneal vascularization, and segmental corneal edema that were steroid-responsive. This clinical picture thus resembled interstitial keratitis, without involvement of the epithelium, and with no other identifiable etiology. It is notable in this respect that the analogous Parry–Robson syndrome also includes ocular inflammatory features such as Fuchs’ heterochromic iridocyclitis and even retinal vasculitis,30 a fact that may imply proinflammatory ocular effects of sympathetic denervation. Neuropathic inflammation may occur after peripheral nerve injury. It is thought to be generated by release of inflammatory mediators such as eicosinoids from injured nerves. The culprit cells in this type of inflammation are assumed to be macrophages and postganglionic sympathetic neurons.31 Thus, Horner syndrome could theoretically induce inflammation in a similar neuropathic mechanism. To our knowledge, there are no previous reports of such clinical manifestations in Horner syndrome. It seems reasonable to assume that sympathetic ocular innervation has more physiologic effects than just lid elevation and mydriasis. Specifically, sympathetic corneal innervation may have a trophic role in normal endothelial physiology that is as yet undetermined. This possibility should, in our opinion, be explored in further studies.

References 1. Thompson HS. Diagnosing Horner’s syndrome. Trans Am Acad Ophthalmol Otolaryngol 1977;83:840 –2. 2. Nielsen PJ. The central corneal thickness in patients with Horner’s syndrome. Acta Ophthalmol 1983;61:467–73. 3. Klyce SD, Beuerman RW, Crosson CE. Alteration of corneal epithelial ion transport by sympathectomy. Invest Ophthalmol Vis Sci 1985;26:434 – 42. 4. Jones MA, Marfurt CF. Peptidergic innervation of the rat cornea. Exp Eye Res 1988;66:421–35. 5. Jones MA, Marfurt CF. Sympathetic stimulation of corneal epithelial proliferation in wounded and nonwounded rat eyes. Invest Ophthalmol Vis Sci 1996;37:2535– 47. 6. Perez E, Lopez–Briones LG, Gallar J, Belmonte C. Effects of chronic sympathetic stimulation on corneal wound healing. Invest Ophthalmol Vis Sci 1987;28:221– 4. 7. Garcia–Hirschfeld J, Lopez–Briones LG, Belmonte C. Neurotrophic influences on corneal epithelial cells. Exp Eye Res 1994;59:597– 605. 8. Zadunaisky JA, Lande MA, Chalfie M, Neufeld AH. Ion pumps in the cornea and their stimulation by epinephrine and cyclic-AMP. Exp Eye Res 1973;15:577– 84.

9. Elena PP, Kosina–Boix M, Moulin G, Lapalus P. Autoradiographic localization of beta-adrenergic receptors in rabbit eye. Invest Ophthalmol Vis Sci 1987;28:1436 – 41. 10. Walkenbach RJ, Ye GS, Reinach PS, Boney F. Alpha 1-adrenoceptors in the corneal endothelium. Exp Eye Res 1992; 55:443–50. 11. Walkenbach RJ, LeGrand RD. Adenylate cyclase activity in bovine and human corneal endothelium. Invest Ophthalmol Vis Sci 1982;22:120 – 4. 12. Gonzalez GA, Feldman ST. Cyclic AMP mediated gene expression in bovine corneal endothelial cells. Invest Ophthalmol Vis Sci 1993;34:2970 –5. 13. Riley MV, Winkler BS, Starnes CA, et al. Regulation of corneal endothelial barrier function by adenosine, cyclic AMP, and protein kinases. Invest Ophthalmol Vis Sci 1998; 39:2076 – 84. 14. Toivanen M, Tervo T, Partanen M, et al. Histochemical demonstration of adrenergic nerves in the stroma of human cornea. Invest Ophthalmol Vis Sci 1987;28:398 – 400. 15. Marfurt CF, Ellis LC. Immunohistochemical localization of tyrosine hydroxylase in corneal nerves. J Comp Neurol 1993; 336:517–31. 16. Garcher C, Humbert P, Bron A, et al. Neuropathie optique et syndrome de Parry–Romberg. A propos d’un cas. J Fr Ophtalmol 1990;13:557– 61. 17. Godfrey WA, Hunkeler JD. Chronic Iridocyclitis. In: Tasman W, editor; Jaeger EA, assistant editor. Duane’s Ophthalmology on CD-ROM, 1996 ed. Version 2.0 for The Windows and Macintosh environments. Hagerstown, MD: Lippincott– Raven, 1996, C. 1995. 18. La Hey E, Baarsma GS. Fuchs’ heterochromic cyclitis and retinal vascular abnormalities in progressive hemifacial atrophy. Eye 1993;7:426 – 8. 19. Miller MT, Spencer MA. Progressive hemifacial atrophy. A natural history study. Trans Am Ophthalmol Soc 1995;93: 203–15; discussion 215–7. 20. Resende LA, Dal Pai V, Alves A. Experimental study of progressive facial hemiatrophy: effects of cervical sympathectomy in animals. Rev Neurol 1991;147:609 –11. 21. Grayson M, Pieroni D. Progressive facial hemiatrophy with bullous and band-shaped keratopathy. Am J Ophthalmol 1970; 70:42– 4. 22. Ford JG, Busbec B, Reed JW, Yu D. Hemifacial atrophy and primary corneal endothelial failure. Arch Ophthalmol 1998; 116:1246 – 8. 23. Pareja JA, Espejo J, Trigo M, Sjaastad O. Congenital (?) Horner’s syndrome and ipsilateral headache. Funct Neurol 1997;12:123–31. 24. Bolden DA, Sternini C, Kruger L. GAP-43 mRNA and calcitonin gene-related peptide mRNA expression in sensory neurons are increased following sympathectomy. Brain Res Bull 1997;42:39 –50. 25. Terenghi G, Zhang SQ, Unger WG, Polak JM. Morphological changes of sensory CGRP-immunoreactive and sympathetic nerves in peripheral tissues following chronic denervation. Histochemistry 1986;86:89 –95. 26. Unger WG, Terenghi G, Ghatei MA, et al. Calcitonin generelated polypeptide as a mediator of the neurogenic ocular injury response. J Ocul Pharmacol 1985;1:189 –99. 27. Unger WG, Terenghi G, Zhang SQ, Polak JM. Alteration in the histochemical presence of tyrosine hydroxylase and CGRP-immunoreactivities in the eye following chronic sympathetic or sensory denervation. Curr Eye Res 1988;7:761–9. 28. Abelli L, Geppetti P, Maggi CA. Relative contribution of sympathetic and sensory nerves to thermal nociception and tissue trophism in rats. Neuroscience 1993;57:739 – 45.

1695

Ophthalmology Volume 106, Number 9, September 1999 29. Gilbard JP, Rossi SR. Tear film and ocular surface changes in a rabbit model of neurotrophic keratitis. Ophthalmology 1990; 97:308 –12. 30. Ong K, Billson FA, Pathirana DS, Clifton–Bligh P. A case of progressive hemifacial atrophy with uveitis and retinal vasculitis. Aust N Z J Ophthalmol 1991;19:295– 8. 31. Tracey DJ, Walker JS. Pain due to nerve damage: are inflammatory mediators involved? Inflamm Res 1995;44:407–11. 32. Clatworthy AL, Illich PA, Castro GA, Walters ET. Role of peri-axonal inflammation in the development of thermal hyperalgesia and guarding behavior in a rat model of neuropathic pain. Neurosci Lett 1995;184:5– 8.

1696

33. Moriwaki K, Yuge O, Tanaka H, et al. Neuropathic pain and prolonged regional inflammation as two distinct symptomatological components in complex regional pain syndrome with patchy osteoporosis—a pilot study. Pain 1997; 72:277– 82. 34. McQuay HJ. Pharmacological treatment of neuralgic and neuropathic pain. Cancer Surv 1988;7:141–59. 35. Boczek–Funcke A, Dembowsky K, Habler HJ, et al. Spontaneous activity, conduction velocity and segmental origin of different classes of thoracic preganglionic neurons projecting into the cat cervical sympathetic trunk. J Auton Nerv Syst 1993;43:189 –200.