OPHTHALMIC HERPES ZOSTER RONALD J. MARSH and MATTHEW COOPER London SUMMARY Fig. 1 shows the age and sex distribution, which is A current review of ophthalmic zoster is presented biased in favour of females and compares with 50.7% including its virology, immunology, epidemiology and males, 49.3% females in another series.5 The 1981 census pathogenesis. We give our findings in 1356 patients for Greater London recorded 48% males and 52% referred to the Zoster Clinic at Moorfields Ey e Hospital, females. London. The treatment of the disease and its ocular com­ ONSET plications is discussed. There is a prodromal influenza-like illness of varying Ophthalmic herpes zoster is a disease varying in severity duration, with headache, pyrexia, malaise, depression, and from devastating, threatening life and sight, to so mild that sometimes neck stiffness, which may last up to a week it may pass unnoticed. The ophthalmic division of the fifth before the rash appears. This is shortly followed by local­ cranial nerve is affected in 7-17.5% of herpes zoster ised pain over the distribution of the ophthalmic nerve, patients. 1-5 Ocular involvement complicates approxi­ lymph node swelling in the corresponding drainage areas mately 50% of these cases and very rarely cases of maxil­ and, occasionally, a red eye. The localised pain is well lary herpes zoster,l affecting many of the tissues of the known to precede the rash by several days in some cases. globe and orbit by highly varied types of lesions. This probably represents the replication and migration We felt it would be helpful to report our experience with phase of the disease and is possibly accompanied by a lim­ the disease because the large number of cases we have ited viraemia. seen has led us to fonn slightly different ideas from many previous publications as to the nature of the disease, its RASH complications and management. We gained our experi­ The rash varies enonnously in distribution, density and ence in the Zoster Clinic which was started at Moorfields severity. It commences as macules which rapidly progress Eye Hospital, London, in 1967 by Professor Barrie Jones to papules, vesicles and pustules. Crusts start to fonn from as partof the ExternalDisease Clinic. Since then the clinic about 6 days onwards. All, or just one, of the cutaneous has expanded and chiefly sees patients referred from branches of the ophthalmic nerve are affected. The lesions Casualty. These patients come mainly from the Greater vary from small, discrete, scattered and superficial to London area and arereferred rapidly by their general prac­ large, confluent and deep with haemorrhagic bullae. The titioners after onset of the disease. A relatively small latter are probably due to a vasculitis in the dennal papil­ number of patients was referred for second opinion. Since lae leading to severe tissue ischaemia. In our patients the 197 1 one of the authors has supervised the clinic continu­ rash was mild in 430, moderate in 743 and severe in 13 1. ously. From 1972 to 1988 all new patients were entered into a specially designed database which was continu­ The average ages for the different degrees of severity of 64 61 ously upgraded. All those with inadequate details or fol­ rash were: years for severe, years formoderate and 56 low-up of less than a year were removed from the years for mild. Oedema database, leaving 1356 patients. The vast majority of is a variable complication, tending to develop patients received no systemic antiviral or steroid therapy after the first 2 or 3 days. It may be so pronounced as to before they saw us and were physically well before the completely close the lids of the affected eye and spread disease started. The figures given throughout this paper on across the midline to involve the other lids (giving the complication incidence are based on this database. How­ erroneous impression that it is a bilateral disease). ever, the series as a whole was slightly biased because Furthennore, oedema is not due to secondary infection in those patients with insufficient follow-up were excluded the majority of cases, since it rapidly resolves without any and most of them tended to have very mild zoster. antibiotic therapy.. �?rrespondence to: R. J. Marsh, MB.BS, FRCS, Department of Differential Diagnosis Chmcal Ophthalmology, Moorfields Eye Hospital, City Road, London ECIV 2PD, UK. The rash can be mimicked by zosterifonn herpes simplex Eye (1993) 7, 350-370 OPHTHALMIC HERPES ZOSTER 351 250 from or developed serious systemic illness. On the other hand, if a child is from a community in which AIDS is [2J Male 200 D Female endemic we would agree that screening should be done. ill 150 OCULAR INVOLVEMENT .0 E Ocular complications can be categorised primarily into ::::l Z 100 those associated with inflammatory changes, those result­ ing from nerve damage, and those secondary to tissue scarring. Inflammatory changes may be in the form of den­ dritic, nummular and disciform keratitis or as a vasculitis in episcleritis/scleritis, iritis, ischaemic papillitis and orbi­ 0-9 20,29 40-49 60-69 80-89 tal vasculitis. Changes resulting from nerve damage 10-19 30-39 50-59 70-79 90+ include neuroparalytic keratitis, some ocular motor pal­ Age group sies and neuralgia. Changes subsequent to tissue scarring are lid deformities, neuralgia and lipid keratopathy. Fig. 1. Age and sex distribution of ophthalmic zoster patients. The course of the ocular disease falls into three phases: which takes on a dermatome distribution.6 Herpes simplex acute, chronic and relapsing. Acute lesions of the globe vesicles are smaller and fr�quently recurrent; they do not and orbit develop within 3 weeks of the rash. They may form the large distinct crusts or the typical punched-out resolve rapidly and completely but can lead to a chronic scarring of herpes zoster and are not as painful. The two course, especially if untreated, and may linger for years. infections may be differentiated by culturing vesicle fluid Alternatively acute lesions may appear to clear but then and assessing fluorescent antibody membrane antigen relapse years after the disease onset - often on suddenly (FAMA).7 Differentiation from impetigo is usually stopping or reducing the topical steroid treatment. Recur­ straightforward because of the lack of dermatome distri­ rence is a particularly distinctive feature of the disease. bution. Very occasionally a brief and mild zosteriform Adequate treatment delivered at the start of the acute rash appears aftertrauma to the forehead or eye. It clearsin phase can significantly reduce severe late and chronic days with no lasting complications. We have no viral cul­ complications. tures on these cases but wonder whether they are caused The old rule that cutaneous involvement of t4e nasocili­ by a closely related or attenuated strain of varicella/zoster. ary nerve heralds ocular complications is a good one (chi­ squared p<O.OI) but not infallible. We found 6 of our 604 SYSTEMIC INVOLVEMENT patients with nasociliary nerve involvement had no ocular Fortunately the vast majority of patients seen by ophthal­ involvement at all. Ve sicles appearing on the lid margins mologists are otherwise healthy, except for those in are almost invariably associated with ocular involvement centres specialising in tumours and immunosuppression; (chi-squared p<O.OI), although it must be emphasised only 12 in a continuous series of a 1000 of our cases had that severe ocular complications may occur with a very malignant disease.8 mild insignificant rash anywhere on the forehead. A small number of patients attending eye clinics develop a systemic vesicular rash and severe illness 1-2 Acute Lesions weeks after the disease onset. Most of these patients tum Eyelids out to have reticuloses, other malignant tumours,9 diseases rash 926 causing immunosuppression such as AIDS, or are iatro­ The lid margin was involved by the in cases. Ptosis genically immunosuppressed (symptomatic zoster). is common and is usually due to mechanical factors amm Furthermore herpes zoster is more frequent and severe in such as infl ation and oedema. Less frequently it is Haemorrhagic bullae patients with these diseases. It is interesting that viruses in neurological. here are a bad sign, the same family as varicella/zoster, the so-called latent heralding severe scarring and all its consequences and viruses (herpes simplex virus, cytomegalovirus and post-herpetic neuralgia. Epstein-Barr) produce severe infections under the same Conjunctivae circumstances.lO Recently an increased incidence of oph­ thalmic zoster has been described in pre-AIDS patients in Catarrhal conjunctivitis is one of the commonest mani­ New York and Africa.11•12 fe stations of herpes zoster, occurring in 1015 patients, and All patients with a systemic rash should therefore be is nearly always associated with vesicles on the lid margin. screened by a clinical immunologist or oncologist for It is generally transitory, resolving within a week, and malignant disease and immunosuppression. We do not rarely becomes chronic. consider it necessary to investigate uncomplicated cases Episclera and sclera of ophthalmic zoster. 8.13 There has been a tradition, largely unsubstantiated, that all young children with zoster should Episcleritis and scleritis are common complications, be investigated for systemic disease. 14 Our series included occurring mildly in 545 and moderately in 208 patients 17 patients under 16 years old none of whom suffered (Fig. 2). Sectoral or diffuse episcleritis usually appears at 352 R. J. MARSH AND M. COOPER 600 membrane which are surrounded by haloes of stromal haze (Fig. 5b). These appear 10 days or so after the onset 500 o Episc - scattered of the disease and are at first white but later become o Episc - nodular brown. Sometimes they underlie preceding epithelial 400 (f) o Intis lesions, but more often they are seen in close proximity to Q; O .0 thickened corneal nerves.l The haloes surrounding them E 300 ::J vary in size and density, are often very sensitive to topical Z 200 steroid, but have a strong propensity to become chronic or to relapse.