Epidemiology

Initially reported in Japan, ARN has now been widely reported throughout Europe and North America, with sporadic cases reported around the world. ARN can occur in patients of either sex and at any age, although there may be a slight male preponderance. Although the disease is typically thought to affect young adults, ARN has been reported in children ^, and in elderly patients as well. Although this disease was initially described in immunocompetent patients, ARN has now been documented in patients with immunosuppression, such as those with AIDS, ^, and in patients on immunosuppression or following bone marrow transplantation.

Clinical features

The disease may affect one or both eyes; most cases begin with unilateral disease. In almost one-third of patients the second eye becomes involved, usually within 1–6 weeks; however, disease in the second eye has been reported to occur up to 20 years after that in the first eye. ^, Most patients have a history of pain, redness, floaters, and blurred vision. An anterior uveitis with or without keratic precipitates often occurs early in the disease, and although a plasmoid aqueous has been described in these patients, hypopyon is rare. Occasionally there is an associated herpes infection at another site, ^, but most patients are healthy, immunocompetent, and without systemic symptoms.

With the onset of vitritis, patients report floaters and diminished visual acuity. Vitritis tends to worsen as cellular immunity to the virus occurs, and the infiltrating inflammatory cells are predominantly lymphocytes and plasma cells. The earliest retinal lesions are small, patchy, white-yellow areas that tend to enlarge, increase in number, and coalesce over time ( Fig. 12-1 ). They usually start in the midperiphery; occasionally they occur in the posterior pole, but do not follow the architecture of the retinal vessels. After several weeks the lesions begin to resolve, with areas of clearing forming a Swiss cheese pattern. At this time perturbation of the retinal pigment epithelium (RPE) also develops. Retinal vasculitis is another manifestation of ARN ( Fig. 12-2 ). Patients often have a severe retinal arteritis. Capillary nonperfusion can be documented on fluorescein angiography, and periphlebitis and venous occlusions have been described less commonly. Retinal hemorrhage can occur, especially in patients with venous occlusive disease. In contrast to findings with other venous occlusive diseases, neovascularization of the iris or retina is uncommon but reported. ^, The choroidal vasculature is also involved in this disease, and fluorescein angiography has shown areas of early hypofluorescence and late staining consistent with ischemia-induced inflammatory changes.

Acute retinal necrosis with dense white retinal infiltrate in retinal periphery.

Posterior pole of the retina 1 week after the onset of symptoms of acute retinal necrosis and after vitrectomy had been performed demonstrating optic disc swelling and retinal perivascular sheathing.

Optic neuropathy also occurs in patients with ARN. Disc edema is a common finding early in the course of the disease, and Sergott and colleagues described two patients with intraorbital optic nerve enlargement. Optic nerve involvement should be suspected in patients with an afferent pupillary defect and severe visual loss with few retinal findings. Visual field testing and color vision assessment can be instructive. Visual loss may be caused by vascular occlusive disease, viral infiltration of the nerve, or optic nerve distension. Optic nerve sheath fenestration has been proposed as a possible therapeutic intervention, but its effect has been unproven in well-controlled clinical trials. ARN has also been reported in patients with clinical findings of viral meningitis.

The acute inflammatory disease tends to resolve over several months with or without therapy. In untreated patients the acute phase usually resolves in 2–3 months. The course of the disease can also be predicted by the number of clock hours of initial retinal involvement. Unfortunately, within the next several months tractional and rhegmatogenous retinal detachments with many large breaks develop in up to 86% of patients. ^, Despite control of the active viral replication in the retina, cellular infiltration into the vitreous and the formation of vitreal membranes composed of RPE and fibroblasts develop and contribute to the tearing and detachment of an already thinned retina.

The Executive Committee of the American Uveitis Society published a set of standard diagnostic criteria for the ARN syndrome, stating that the designation of ARN syndrome should be based solely on clinical appearance and the course of infection. Clinical characteristics that must be seen include (1) one or more foci of retinal necrosis with discrete borders in the peripheral retina; (2) rapid progression of disease if antiviral therapy has not been given; (3) circumferential spread of disease; (4) evidence of occlusive vasculopathy with arteriolar involvement; and (5) a prominent inflammatory reaction in the vitreous and anterior chamber. Optic nerve involvement, scleritis, and pain support but are not required for the diagnosis, and macular lesions, although less common, do not preclude a diagnosis of ARN if peripheral lesions are present.

Etiology

ARN was initially believed to be an autoimmune disease. In 1982 Culbertson and colleagues described the first occurrence of an enucleated eye in a patient with ARN, in which herpes virus particles were noted in the retina on electron microscopy. However, virus was not cultured from the eye, and serum antibodies were found to herpes simplex, herpes zoster, cytomegalovirus (CMV), and EBV. Histologic examination showed prominent retinal arteritis and marked retinal necrosis with abrupt demarcation between normal and necrotic retina. Optic nerve inflammation was also noted. A similar histologic picture was seen in a retinal biopsy specimen we obtained from a patient with ARN in 1983, which revealed inflammatory cell infiltration and necrosis of the retina ( Fig. 12-3 ). There was a predominance of T lymphocytes in the vitreous, but the retina contained many B lymphocytes, which suggested that antibody to the virus could be produced locally in the retina. Finally, the results of electron microscopy showed herpes class viral particles ( Fig. 12-4 ), although virus was not cultured from the specimen. In 1986 Culbertson and colleagues used immunohistochemical staining to reveal herpes zoster in two eyes with ARN, and also succeeded in culturing the virus from one of the two eyes. Varicella-zoster virus was initially thought to account for most of the typical cases of ARN; however, herpes simplex types 1 and 2, and rarely CMV were also implicated as a cause of the disease. Herpes simplex virus was also implicated in ARN because of serially increasing serum and intraocular antibody levels; however, these may result from a nonspecific polyclonal activation of B cells in the inflammatory process rather than a humoral immune response against a specific virus involved in the current disease.

Histologic section of chorioretinal biopsy specimen from a patient with acute retinal necrosis showing lymphocytic infiltration and areas of necrosis. (Hematoxylin & eosin stain; ×400).

(Courtesy of Chi-Chao Chan, MD.)

Electron micrograph demonstrating herpes class virions in retinal biopsy specimen from a patient with acute retinal necrosis.

In a retrospective study of 28 patients, Ganatra and colleagues used polymerase chain reaction (PCR) to show that varicella-zoster virus or herpes simplex virus type 1 caused ARN in patients older than 25 years, whereas herpes simplex virus type 2 caused ARN in patients younger than 25 years. Van Gelder and associates also report that herpes simplex virus type 2 is an important cause of ARN, particularly in young patients. Sugita and colleagues more recently analyzed ocular fluid samples from 16 patients with ARN. High copy numbers of HSV1, HSV2, or VZV were found in all the samples from the 16 patients, suggesting active viral replication.

Although data from PCR support the role of herpes virus infection in ARN, it is difficult to recommend the use of PCR for the diagnosis of ARN. First, vitreous specimens are probably needed for the analysis. Second, some caution in interpreting PCR data is warranted. PCR is so sensitive a test that viral DNA from a previous infection years earlier may yield confusing results, especially because viral infection with herpes simplex, CMV, and herpes zoster are quite common.

Other experimental data support the role of herpes virus infection in the pathogenesis of ARN. Injection of herpes simplex virus type 1 in BALB/c mice produces a necrotizing retinitis in the contralateral eye within about 10 days. This animal model is similar to ARN in humans and has yielded interesting immunologic insights into the pathophysiologic course of the disease. The necrotizing retinitis develops only in immunocompetent animals. In addition, disease develops in the contralateral eye only if it is infected with virus via the optic nerve. Thus, as in other inflammatory eye diseases, retinal damage, initially induced by an infectious agent, leads to a secondary immune response against previously sequestered retinal antigens. This secondary immune response can then propagate the inflammatory damage to the retina.

Other factors may play a role in the pathogenesis of ARN. Holland and colleagues showed an association between ARN and HLA-DQw7 and phenotype Bw62, DR4. These data suggest that some persons may have a genetically determined predisposition to mounting an immune response against certain infectious agents that promotes the development of ocular inflammatory disease. Some data provide evidence for antigen-specific immune deviation in patients with acute retinal necrosis. Varicella-zoster virus-associated ARN developed in patients in whom delayed hypersensitivity to viral antigens was absent. In this study, delayed hypersensitivity responses were restored in patients who recovered from the disease. In another study, Fas and Fas ligand expression were absent in a retinal biopsy from a patient with ARN. Furthermore, apoptosis was noted on the specimen, suggesting a role in the disease. Finally, inflammatory cytokines may be involved in the development of ARN. Intraocular inoculation of herpes virus induced vascular endothelial growth factor, flk-1, transforming growth factor β ₂ , and interleukin (IL)-6 in the retina of injected and contralateral eyes in experimental animals.

Differential diagnosis

Because ARN can often begin as an anterior uveitis, examination of the peripheral retina is important in all patients with a new onset of anterior inflammation. The differential diagnosis of ARN is listed in Box 12-1 . In general, only Behçet’s disease and endophthalmitis mimic the rapid progression and severity of ARN. The lesions of Behçet’s disease rarely present with a uniform distribution in the periphery of the retina and are often associated with systemic symptoms and signs such as arthritis, mouth and genital ulcers, and skin lesions. Vitrectomy may be needed to rule out endophthalmitis. Toxoplasmic retinochoroiditis, syphilitic retinitis and intraocular lymphoma more rarely mimic ARN.

Therapy

Many patients appear to benefit from treatment with aciclovir. Aciclovir has good activity against herpes simplex virus and, at slightly higher concentrations, is effective against herpes zoster. Although aciclovir is not as good a drug as ganciclovir or foscarnet for treating CMV infection, because CMV is rarely a cause of ARN and aciclovir is easier to administer and safer, it is the best initial therapy for the disease. Patients with newly diagnosed ARN can be treated with a 10–14-day course of intravenously administered aciclovir 500 mg/m ² every 8 hours. Oral aciclovir 800 mg five times a day is then continued for an additional 6 weeks. The goal of therapy is to hasten the resolution of disease in the infected eye and to prevent contralateral spread. However, the therapeutic benefits of aciclovir for ARN have not been studied in well-controlled, randomized clinical trials. Severe side effects with aciclovir are relatively rare. A reversible rise in the serum creatinine level and elevation in liver function test results may occur. The most commonly reported side effects include nausea, vomiting, and headache, but most of these occur in fewer than 3% of patients.

Unfortunately, there have been no randomized prospective clinical trials of aciclovir specifically for the treatment of ARN. In a retrospective study by Palay and colleagues, 87.1% of patients treated with aciclovir did not develop contralateral disease, compared to 35.1% of patients not receiving the drug. Other case series seem to show that retinal lesions resolve more quickly after initiation of aciclovir, but therapy does not appear to diminish vitritis or to prevent subsequent retinal detachment. ^, Patients with severe and progressive disease despite aciclovir therapy may be treated with ganciclovir or foscarnet. In addition, systemic corticosteroids such as prednisone 0.5–1.0 mg/kg can be added after the patient has received 24–48 hours of intravenous aciclovir. Steroids can be slowly tapered based on the amount of vitritis. Because of the extensive retinal arteritis and development of retinal vascular occlusions, some have suggested aspirin therapy, but again this therapy has not been scrutinized in a clinical trial. Intravitreal injections with antiviral agents may be useful as adjunctive therapy in patients with ARN.

Both valaciclovir and famciclovir offer more convenient dosage regimens. Famciclovir at a dose of 500 mg three times daily had efficacy similar to that of aciclovir in treating patients with ophthalmic zoster. Valaciclovir is rapidly converted to aciclovir and the usual dose is 1 g orally three times daily, resulting in substantial vitreous penetration of aciclovir. Importantly, in a 1-year prospective randomized clinical trial, recurrence of any type of ocular HSV disease was 23.1% in patients receiving valaciclovir compared to 23.1% in patients receiving aciclovir. The use of either oral valaciclovir or famciclovir has resulted in regression of retinitis in patients in several case series.

Retinal detachment and optic neuropathy remain the causes of severe visual loss in this disease. Laser photocoagulation to demarcate a zone between involved and uninvolved retina has been tried as a prophylactic therapy to prevent retinal detachment. Sternberg and colleagues showed that retinal detachments occurred in two of 12 eyes receiving laser therapy and in four of six eyes not receiving laser therapy. However, this retrospective analysis may be biased because the condition of patients with the most severe disease and the worst vitritis may not have been amenable to laser therapy. Others have suggested prophylactic vitrectomy, scleral buckling, and endolaser treatment for eyes at high risk for retinal detachment, but there are few data to support this therapy.

Clearly, surgery is warranted to repair retinal detachments in patients with ARN, although these detachments are difficult to repair and proliferative vitreoretinopathy is common. Vitrectomy, fluid–gas exchange, and endolaser treatment have been suggested for complicated detachments. Whether this therapy is better than vitrectomy with scleral buckling is unknown. Silicone oil has also been used successfully in some patients.

Less well studied is the effect of therapy on optic nerve disease. Optic nerve fenestration and corticosteroids have been used to treat optic neuropathy in ARN, but no well-controlled study data are available to support the effectiveness of these therapeutic interventions.

With the discovery of the viral etiology of ARN, specific therapy has been employed for the treatment of this disease. Nevertheless, many questions about ARN remain. We still do not know why certain individuals are more susceptible to developing the disease, because ARN is relatively rare but herpes virus infection is quite common. Finally, although aciclovir, valaciclovir, and famciclovir and corticosteroids are used to treat patients with ARN, there are still no randomized, masked, clinical trials definitively showing the benefits of therapy.